Loading summary
Transcript115 lines
- [00:01]
UNHCR Narrator
Displacement is at its highest level since World War II, but more than $1 billion in essential programs are being cut just as needs grow. Families forced to flee war are arriving in camps hungry, cold and exhausted. With your support, unhcr, the UN refugee agency, provides essentials for emergencies. Warm clothes, blankets, cooking sets, and shelter materials help families survive the deadly winter ahead. Donate@unrefugees.org Scene.
- [00:40]
Tyler Foggit
Welcome to the political scene from the New Yorker. I'm Tyler Foggit. 2025 was the year that artificial intelligence pushed its way into our politics and into all of our lives, whether we liked it or not. Today we're bringing you a conversation that I had with the writers Charles Duhigg, Cal Newport, and Anna Weiner at the New Yorker Festival in the Flor all we looked at the current state of AI, the kind of world being created by it, and how all of us might adapt or recalibrate to the technology. I hope you enjoy the conversation.
- [01:15]
Moderator
So I guess I thought we'd start by talking about the two competing narratives about technology and AI that we often hear about. So you got the first narrative, which is that, and this is the narrative that's being sold to us by the tech giants, and that's that these tools will make us better. They will make our lives easier, more efficient. AI will eliminate our busy work, and it'll lead to incredible discoveries in science and medicine. And then there's the other narrative, which is that these tools are ruining us. AI is going to replace us and our jobs entirely. It might even kill all of us if it decides to be evil. You know, social media, meanwhile, and smartphone dependency are destroying our attention spans and our ability to think for ourselves. We were told that these tools would connect us, and instead they're leading to a fractured, helpless, sad society. So my question for you guys is which of these narratives feels more correct right now, or do you think it's like a mix of the two? Do you want to start?
- [02:12]
Katie Drummond
Charles?
- [02:14]
Charles Duhigg
Sure. So I think the thing that's really interesting is we've fallen into this pattern where we talk about this as a binary outcome, right? There's this instinct to see this as good or bad. And if you think about most of the technology that's influenced our lives, this is pretty typical historically, that when a technology comes into being, we tend to look at it in a binary way, but then the application of it is incredibly non binary, right? It's much more in between. It's much more of a normal distribution. Telephones is one of my favorite examples. The number of articles that appeared when telephones became popular, saying that it would destroy society, that people would never have a real conversation on a phone, that it would pollute children's minds so much that they became mute and could not communicate face to face. All of which ends up being a little bit true, right?
- [03:07]
Cal Newport
Like, we're not.
- [03:08]
Charles Duhigg
We don't feel great about our kids on our phones today. But I think that for me, the thing that I keep thinking about is there's this instinct to see this as good or bad. And it's much more natural for us to use this thing in a way that it's neither good or bad, it's both.
- [03:26]
Cal Newport
Well, so there's another axis, I think, about this. It could also be useful. So if you're watching the rhetoric coming out of the AI companies and then the rhetoric of people who are talking about the AI companies, I like to think that we have just passed in the last two months from Act 1 of the Gen AI revolution to what I now think of as Act 2. So in Act 1, there is a lot of focus on predictions. You know, hey, this is an amazing technology. Let's react to what it could do. So there's a lot of discussion about, so what will happen if and when it automates half of our jobs? Like, what will happen when this industry, this industry, the industry disappears? So it's a lot of looking at the predictions about what this could do. There was a pretty astonishing shift, I think, in the discourse right around the August release of GPT5, where we moved past saying thought experiments, what this could do, and reacting to that, to, well, what does it do right now? And I think this is where we are now in the last couple of months has been starting to grapple with, well, what specifically do these tools do? Where are they good and where are they bad? Which I think is closer to the way you're talking about it. Charles I think this has been or is going to be a very healthy transition, because up until now it was very difficult to really have a pragmatic conversation about AI because we had superintelligence and job automation. We had Sam Altman talking about how the release of GPT5 was going to be the biggest cultural event since the Trinity test of the first atomic bomb. We had Mark Zuckerberg saying superintelligence was now within sight. He was just making that up. We had Dario Amade saying probably half of new white collar jobs will be gone within three to five years. We were entertaining thought experiments. And I think right now the discourse in tech journalism is now saying, oh, forget that. Like, what exactly are we using these tools? Oh, that's good, that's bad. Oh, here we might want to control that a little bit. So I think we're now in this more pragmatic grappling phase of the new technology curve. I think it's a better phase that we're in anyways. That helps me think about what's been shifting recently.
- [05:25]
Anna Wiener
I really love the way that Sam Altman, Mark Zuckerberg, the way that they talk about this technology, just the prose style is really unique to me. I think that where I fall on all of this actually has less to do with the tech itself and more to do with how the sort of social structures are around it. Both that will kind of facilitate its normalization or its use, and also how well equipped are we to deal with potential repercussions. So I'm a little more interested in that than in. I mean, just to be candid, I don't actively seek out generative AI tools personally. They're interesting to me on a sort of cultural level. But yeah, like, how well equipped are we to absorb the effects of this technology? What needs to change? Who will it benefit? Like who, who. Who will suffer because of this? And. And how do we mitigate that suffering or. Or save it off the past? So those are some of the questions that kind of animate my feeling about it between the marketing and the doomer perspectives.
- [06:22]
Tyler Foggit
Cal. Charles, how often are you guys using generative AI tools? And I'm wondering if you can speak about your personal experience using these things or spurning them.
- [06:31]
Charles Duhigg
I use it probably every other day. And I, and, and I just learned how to vibe code. I went to this thing and they taught me how to vibe code, which just like having a conversation, it's not that different, but I'd say at least two or three or four times a week and sometimes five or six times a day.
- [06:48]
Cal Newport
I mean, I'm mainly using it as an enhanced Google occasionally. I feel like this is what gets me trying more tools. So right now I'm in the middle of something where I'm using an AI tool to help clear my inbox. They're very good at filtering messages, so figuring out this is important or this is not important, but if not for that, it probably wouldn't have a big footprint yet in my life.
- [07:11]
Anna Wiener
Can I ask a question to Charles?
- [07:12]
Tyler Foggit
Yes.
- [07:13]
Anna Wiener
What did you vibe code?
- [07:16]
Charles Duhigg
What do you not want to vibe code? You can vibe code anything.
- [07:20]
Anna Wiener
Did you make something?
- [07:21]
Cal Newport
Yeah, yeah, yeah.
- [07:21]
Charles Duhigg
So I made something that told me if there was a it was using Rep leaf, if anyone uses it. And I said build me an app using this data set that's available online to tell me if there's going to be a thunderstorm in my area. Because I was just trying to figure out what to use it for. And it is cool. But, but this actually gets to a point that I think you raised about the absorption, which is if you think about the last 25 years, last 30 years, the big things that changed our world were not necessarily just advances in technology, right? It was Jeff Bezos realizing that now that the Internet exists, you can use.
- [08:00]
Cal Newport
It to sell everything.
- [08:02]
Charles Duhigg
And that was true for the 15 years before Bezos started doing that. Or it was the guys at Google realizing that page rank that what's more important is figuring out if other people link to a page to start the search engine. And so I actually think that there's a cognitive barrier that no matter how advanced the technology gets, it'll be used by a very small number of people and they might be terrorists creating Ebola viruses. So that's bad. But for most of us, the technological breakthrough is not necessarily going to be the technology. The technological breakthrough is going to be a new concept of how we use that technology and the possibilities of the world. But I'm wondering if you guys disagree.
- [08:40]
Cal Newport
I mean, I think that's right. I think if you look, let's use the search engine example, right? If you look pre Google there was something like two to three dozen search engines that were out there. So they weren't innovating. The concept of this technology exists which can crawl the web and return search results. And the really the primary breakthroughs of Google were infrastructure breakthroughs figuring out how we can use low cost computing components in a very distributed way that can be very fault tolerant and allow us to actually deal with way more data. That there's a computer science story there. But that infrastructure changed the experience of this technology that already existed in a major way. I think social media was similar. We have social media, web based social media for a while. Then Instagram comes along and begins to figure out some of the technology needed to make this native on a phone. Because it was a phone based app, had to be native on the phone. And this was like infrastructure breakthroughs that allowed attention harvesting and attention engineering to occur on an exponentially bigger scale. We started using our phones way more. So I think that's interesting. We're looking now at chatbots that show that generative AI exists and it has a lot of power in processing text and producing text and other types of media. But yeah, maybe we haven't. We don't know yet what the sort of infrastructure breakthroughs, app architecture breakthroughs are that makes that a killer app in a way that you're right. I think we might not be able to predict that yet. We often get that wrong.
- [09:59]
Anna Wiener
I mean, it does seem like there's a sense of inevitability that's already pervading this conversation that we all sort of seem to tacitly agree that this is a technology that is here to stay and will be revolutionary even if we haven't yet reached that point. And I guess I sort of want to challenge that assumption that seems to be underlying the conversation so far. Let's start there, I guess. Do you guys see this as an inevitable massive breakthrough that will transform the next hundred years of, you know.
- [10:24]
Cal Newport
Can I ask you a follow up question on your question? Right. So I guess we should. What technology are we talking about? I guess.
- [10:30]
Anna Wiener
Right. Well, you were talking about generative AI and specifically text.
- [10:36]
Cal Newport
Yeah.
- [10:36]
Anna Wiener
And I was thinking, you know, we haven't really even talked about any the other forms of media generation yet. And you know, we might see that, that maybe image generation or video generation accelerates faster than certain kinds of data processing. You probably have more insight into the actual.
- [11:01]
Cal Newport
I can get my one minute computer science view of what I think is going to happen with the technology, because I think what's happening now is there was this belief, and we did a piece about this in August, there was this belief that you could take something like a text based LLM, like one of the GPT family of models, and that if you trained it on enough information, and you trained it long enough on enough information that just this text producer would end up internalizing enough logic and understanding of the world that it could be generally intelligent. A language model itself, if you trained it big enough, would have something like artificial general intelligence. That's a storyline that within the industry came to an end last year and the public realized this sort of last summer that actually that's not going to work. So text. My theory is there's a lot of work to be done to how do we actually take this idea? We have a tool that can understand text really well and integrate this into all sorts of products and that it's going to be really useful. But my vision of like artificial general intelligence is that I think it's going to be distributed in the sense that we are going to have many, many different types of systems that use many different combinations of different types of models, Some of them using breakthroughs that happen with language models, but some using other types of models, more neuro symbolic AI as well. And it's not going to be one oracle that does everything for us. It's going to be, here's a system that's really good at doing exactly this task. You can do it a little better than people. Here's one that can do this task. It's its own startup, it can do that a little better than people. And you multiply that by a thousand different tasks and then you look around like, oh, I guess we kind of got to artificial general intelligence. Not with like a HAL 2000 that does everything, but just we have figured out task by task how to sort of reconfigure and pull together these ideas to get systems to handle it. The industry does not agree with me on this, by the way, but that's sort of my vision. I don't know if it's going to happen, but that's where I am now.
- [12:50]
Charles Duhigg
So I want to answer your question and I think this ties into what you just said. So if you think about again, going back to telephones, when telephones were invented, what's interesting about the phone is that the scale of invention there, the scale of transformation, was relatively low. It replaced the telegraph. And with the telegraph we could basically do the same thing that's useful in our lives with the telephone. But what it did is it eventually democratized it. Then in comparison, there's the Internet. And I would argue you could actually say the Internet is more powerful than the invention of the telephone, because the Internet has, has slowly but eventually transformed everything. It's transformed how we date, how we buy things, how we decide on what jobs or colleges to attend. I think that AI is closer to the Internet than it is to the telephone. I'll make that argument. I'll invite you and everyone else to disagree with me.
- [13:41]
Cal Newport
You mean language models in particular or AI more broadly?
- [13:44]
Charles Duhigg
I think AI more broadly. AI more broadly. But neural nets. Let's say if this technology proves to be useful and eventually we have hardware robots that can use this technology to move through the world and do things for us. I think it's more akin to the Internet. And I will say, much like the Internet, if in 1995 you tried to forecast how that technology would be transformative, the best thinker on earth would get probably 20% of it. Right? Right. It's, it's moved so much more slowly and so much more differently through our lives than we imagined initially. And so I think that's why it's going to be profound. Do you agree or disagree?
- [14:22]
Cal Newport
Yeah. Where are you?
- [14:23]
Anna Wiener
Seems reasonable. Would you consider it a platform?
- [14:26]
Charles Duhigg
No, I get. I. I think of it more as like a tool, but maybe we're using words differently. What do you mean?
- [14:33]
Anna Wiener
I don't know. I'm just curious. From your description, I mean, I think Cal I sort of land where you are in terms of it being. I think your approach is really pragmatic. It's also obviously informed by your background. I live in San Francisco and the narratives, it's like a very storytelling oriented industry. I think there's a lot of enthusiasm for larger than life omniscient entity. Whereas I see this as more of a sort of an additive technology that like you're saying, will improve certain things probably from the bottom up quite slowly.
- [15:09]
Cal Newport
There's actually an east coast, west coast feud going on among technologists. So like on the west coast you have technologists based mainly in industry who are very utopian about this. And on the east coast you have a lot of sort of dower east coast computer science professors who are, who are looking at the technology and saying like, I don't understand how this is going to do this or that and are being curmudgeonly about it and we're not getting along right now. Sort of interesting.
- [15:33]
Charles Duhigg
When I moved to Northern California from Brooklyn in 2020 and I used to say, and I think this is still a little bit true, that in California all anyone talks about is AI and psychedelics. And in New York all anyone talks about is AI. That's pretty much the difference. And that might explain that divide between the two coasts.
- [15:52]
Anna Wiener
Okay, I've been in New York for 10 days and I've only had one conversation about AI, except for this one, and that was my fault. So I feel like it depends on the company you keep, it can be found.
- [16:05]
Charles Duhigg
That's absolutely true.
- [16:06]
Cal Newport
Yeah. Well, can I throw out a provocation about what you were saying, Charles? So I'm very interested when we have these new technologies that become potentially revolutionary and they completely transform things. What the story is when they arrive, like what is obvious and what's not obvious. Right. So I feel like there might be daylight here between the current crop of AI technologies and the Internet, because when the Internet was coming along, there was a very clear story to be told there, which is anyone can publish information, anyone can read it. Right. So this was a hugely democratizing, but also very easy to understand. Like this is very different. Like never before in History. Could it be because we have shared protocols that everyone uses, anyone can put up information, anyone else can read it. So we were wrong in our predictions of exactly what that was going to lead to. But that felt like this is like printing press in the sense of, oh, now we can produce books 10x faster, 100x faster, whatever. It feels a little murkier still right now with the current AI tools. It feels a little bit murkier about is it can generate text, it can understand text, it can generate video, but it doesn't yet have that self evident feel yet. I talk about email with something similar. Like when email came along it was like self evidently useful to everyone. It was like, this is great, this is a much more efficient way to send information. And I feel like AI hasn't, it's still new enough and we don't have the killer apps yet. And so there's something different. There's like a bit of a different valence that it feels like, yeah, artificial intelligence as a concept should be revolutionary, but it's a little bit harder to put our finger on unless we go into just pure thought experiments. Like, well, it can automate whatever anyone is doing. But it feels like this is what's happening now as we're trying to figure out, well, what exactly is going to be the key drivers of the near future revolution of this.
- [17:56]
Charles Duhigg
So I completely agree and I think we are in this very plastic moment, much more so than at the birth of the Internet, with one exception, which is the investment that's happening right now in AI, that there's this absolute certainty about how much should be invested. So just to put this in scale, because particularly in New York, I think you see less of it. The Apollo program putting people into space took 10 years and in today's dollars it cost about $400 billion. The AI industry is spending $500 billion a year building out infrastructure. So we're doing the equivalent of 10 years of Apollo spending basically every 10 months. And there's a real legitimate question, is that a bubble? Is this going to be all for? Not is it wasted money, is it real money? But the truth is that we are building out an infrastructure that will be, I think, transformative regardless of what the transformation looks like or what it's used for. Just because we're building so much capacity.
- [18:55]
Cal Newport
It's like fiber in the late 90s.
- [18:56]
Charles Duhigg
Exactly.
- [18:57]
Cal Newport
Which a lot of those companies went out of business. But then we use all the fiber today.
- [19:03]
Tyler Foggit
The political scene live at the New Yorker Festival. We'll be back in Just a moment.
- [19:16]
Katie Drummond
What the hell is going on right now and why is it happening like this? At Wired, we're obsessed with getting to the bottom of those questions on a daily basis. And maybe you are too. I'm Katie Drummond, the global Editorial Director of Wired, and I'm hosting our new podcast series, the Big Interview. Each week I'll sit down with some of the most interesting, provocative and influential people who are shaping our right now. Big Interview. Conversations are fun.
- [19:43]
Jason Adam Katzenstein
I want a shark that that eats.
- [19:45]
Katie Drummond
The Internet, that turns it all off, unfiltered and unafraid.
- [19:49]
Cal Newport
So in a lot of ways, I try to be an antidote to the unimaginable faucet of reactionary content that you see online. To the best of my ability, every.
- [19:58]
Katie Drummond
Week, we're going to offer you the ultimate luxury of our times. Meaning and context. True or false. You, Brian Johnson, the man sitting across from me. One day, at some point, as of yet undefined in the future, you will die. False. Tell me more. Listen to the Big Interview right now in the same place you find WIRED's Uncanny Valley podcast. Subscribe or follow wherever you get your podcasts.
- [20:27]
Jason Adam Katzenstein
I'm New Yorker cartoonist and perennially cold man, Jason Adam Katzenstein. Cold mornings Holiday plans this is when I need my wardrobe to just work. Which is why I'm looking at the Mongolian cashmere puffer coat from Quince. Quince pieces are crafted from premium materials and built to hold up without the luxury markup, which is great because I don't want to break the bank, but I also don't want to be shivering anymore. Quince makes essentials that I need. Their outerwear lineup's no joke. Down jackets, wool topcoats and leather styles that are built to last. Each piece is made from premium materials by trusted factories that meet rigorous standards for craftsmanship and ethical production. It's everything that I actually want to wear. Built to hold up season after season so that I can hold up season after season. Get your wardrobe sorted and your gift list handled with quints. Don't wait. Go to quints.com politicalscene for free shipping on your order and 365 day returns. Now available in Canada too. That's quince.com politicalscene free shipping and 365 day returns. Quince.com politicalscene.
- [21:42]
Tyler Foggit
Do you think there's ever a point at which these companies who are sinking tons and tons of money into this project that they would ever admit that we're just never going to have a super intelligence? Or is that kind of Impossible, given how much money is going into this.
- [21:56]
Charles Duhigg
Well, I think what they would quibble with is what's the definition of superintelligence? So, like, you guys have seen Meta say like they're building superintelligence, which is like, even everyone in the Valley is like, that is such bs. Like, like it's basically a marketing message that you want to attract people to, to work on cool problems. But so I think that there's an interesting question. The problem is we use these phrases like superintelligence or AGI, artificial general intelligence without actually def. What they are. And the best way to do it, I think, is to root it in analogies of things that we are familiar with. Like, does a dog have general intelligence? I would argue yes, but other people could argue no. And if so, the things that make that dog intelligent, can a machine replicate that in a way that doesn't particularly concern us? So I don't think that companies are actually having that debate as much as the debate that they are having is for the last 15 years, everything in tech has been winner take all. And if we don't invest $500 billion, we won't be the winner and someone else will and we'll go out of business.
- [23:02]
Tyler Foggit
Yeah, I mean, I guess it's.
- [23:03]
Moderator
It's very hard to define what a super intelligence even is, but I kind of feel like I know what it's not and that's what we're getting right now, which is like, I try to use, you know, chatgpt for very specific kind of Google search, and sometimes it can be very good at that, and sometimes it gives me just like a totally just hallucinates. And even when I'm asking it something very simple. And so I'm like, am I doing something wrong? Or is this just like a fundamental problem? And like, why is it so hard?
- [23:31]
Charles Duhigg
Or if you use Alexa, it's super unintelligent.
- [23:34]
Jason Adam Katzenstein
Right?
- [23:34]
Charles Duhigg
Like, it's like Alexa is like, I'm not worried about the robots.
- [23:36]
Tyler Foggit
Yeah.
- [23:37]
Cal Newport
I mean, the entire. The story of superintelligence, I think, is nonsense. The entire technical story for superintelligence is this idea of takeoff. So it's this idea of recursive self improvement that once we make a machine smart enough to improve itself, then that next machine will be smarter and it will improve itself to be even smarter, and then that machine will improve itself to be even smarter. So this rapid takeoff scenario, they call it, that's at the root of the superintelligence thought experiment, which is we have an AI so smart that we don't understand it and then it can conquer us. This is the new book. If we build this, we'll all die. That's the concept from a computer science standpoint that recursive self improvement language models can't do that. It's just nonsense. Right? I mean the, they cannot solve the problem of producing a program that has like a brand new theory of intelligence and how to build this into a computer program. I think that's nonsense. But you will see that the rhetoric has changed if you can put a timeline on the CEO's rhetoric. 2023 Early ChatGPT it was a lot of superintelligence rhetoric. It's like we got to be worried the whole world's going to change. Sam Altman saying we're going to have to have an equity tax because there'll just be two companies left and everyone else we'll have to have a way to eat. And they really believe that. Then they all shifted at some point they're like, oh, this is scaring investors so en masse. At some point in 2024, everyone started talking about job automation. Now this is what the power is going to be. It's not super intelligence. It's going to automate jobs and we have to grapple with this in the economy. And that was a better story for investors because if your company is automating jobs, your company is going to be worth a lot of money. I want to invest in you. And now they've shifted again because the year of the Agents never happened. We have the new MIT study that found 95% of the deployments they studied of AI in corporate environments had zero return. And so now they've shifted the rhetoric again. So I feel like there's, I mean, you know these leaders, you probably have a better sense of them than I do, but there's like a sort of coordinated shifting that keeps happening and they want there to be something to emphasize that makes it seem like this is an inevitable, powerful enough technology you should invest. But they keep sort of shifting what that target is and they sort of like semi coordinated. Now they're talking about bubbles and about. So it's interesting, like they keep shifting the way they talk about. So they have talk about super intelligence as much as they used to.
- [25:47]
Charles Duhigg
I was just going to ask you a question actually, because I. You write very beautifully about like the importance of human connection and how people connect to each other. I felt like in, in Uncanny Valley and you've been talking to people, I think, who are using AI for connection. We've been talking about very practical applications, but do you think that there might be a transformation in how people feel about themselves in the world because of a relationship they develop with AI?
- [26:14]
Anna Wiener
Yeah, I do. I think I want to just build on something that Cal was saying and then I can address that, which is just these discursive shifts. And I'm old enough to remember when we were still calling it machine learning. These terms get normalized and these concepts get normalized, and they're great stories. Just to go back to Tyler's first question, and people love a good story. And I think Silicon Valley is ham fisted in a thousand ways, many of them aesthetic, and yet they've managed to be very good storytellers about themselves. And so the sort of adoption of something like ChatGPT and the speed with which people are suddenly realizing that they need to care about this, that to me is kind of phenomenal. You can't open the New York Times without reading a sort of anxious op ed about AI in the future and all of the uncertainties. I just want to highlight that because I do feel like they're speaking to investors, but they're also speaking to the public. What's interesting to me, it's funny that earlier we were talking about how we use ChatGPT, stuff like that. And I think too, Tyler, maybe you mentioned using it for research, but then you're like, it hallucinates all the time. It's like having a very bad Internet where you're like, well, I have to check this work anyway, so I guess I'll just do it myself. And yet it's still in the mix, significantly enough that someone like you is still using it, even though you probably know how to do research and are incredibly good at finding the information that you need. So I don't know, that's kind of an interesting thing to me, I think in terms of the emotional relationships that people are developing with LLMs and AI companions and different products that are marketed to people who. And the people who use them. I think that the emotional relationships that people have with this technology, I mean, part of it is because something like ChatGPT or Claude, they're quite manipulative, I would say, in the way that they speak and the way that they sort of train the user to address them. I think it's very, very strange that people interact with something like chatgpt politely and maybe that speaks to the goodness of humans. But it's the software telling us how to talk to it. We're being trained to use it in A certain way. I know that sounds a little tinfoil hat, but there's a reason that it looks like a text message thread or an email thread. They go along. I feel actually morally neutral personally about the use of LLMs in products that are meant to be emotional because I have talked to so many people who find great value in them and who are aware that what they're engaging with is an app and not a human. I think there are all sorts of things happening in people's lives that might lead them to engage with something like that, but the feelings are real. And so I think in evaluating that, it's really hard. You can't really place a judgment on that. But what I am finding is that the feelings are very real. And. And it's sort of. That, to me is a story that runs alongside why people are. Are being drawn to these tools in the first place. Sorry, that was quite a yappy answer.
- [29:36]
Charles Duhigg
No, it's super interesting.
- [29:37]
Anna Wiener
That was the Chat GPT answer.
- [29:40]
Charles Duhigg
That was a good.
- [29:41]
Anna Wiener
Not enough bullet points.
- [29:42]
Charles Duhigg
Great response. The best response I've ever heard.
- [29:45]
Cal Newport
You are the best responder in the world, Anna. Here's an emoji.
- [29:49]
Charles Duhigg
Internalize the Chat GP a little bit.
- [29:53]
Anna Wiener
No, it's.
- [29:54]
Tyler Foggit
I mean, so there are these very.
- [29:55]
Moderator
Extreme examples of like, people who feel like they're married to, you know, the AI, but like, what about the, I.
- [30:02]
Tyler Foggit
Guess, like milder ones where it's like someone who's using ChatGPT as a therapist or as a debate partner, who just happens to know like, every single possible statistic and will never verge into like, you know, like ad hominem and like, the kind of things that make humans bad debate partners sometimes? Like, do you think that there are ways where we can farm out human relationships to ChatGPT in a way that isn't kind of weird and toxic?
- [30:28]
Charles Duhigg
So I was telling these guys backstage, I have a 17 year old at home who's a smart kid and he's applying to college. And so this summer we were helping him work on his essays and he really, really wants to go to Brown. So if there's anyone from the Brown admissions committee in the audience. But. So we were working on his essays with him and then I was looking at my Chat GPT history because he uses my account. And one of the queries was, I want to go to Brown. This is my essay. Do you think I'll get in? And of course, Chad GPT was like, not only are you going to get in, you're going to be one of the top choices to get in. This is amazing. So I went and I like kind of talked to Ali about this and asked him like, are you also asking it like, you know, if you should commit suicide or anything like that? I just want to make sure we're like, like nothing bad is happening here. And he was like, no, no. I just sort of felt, I felt a little bit down about it and I knew that it would make me feel better, which I think is like kind of interesting. To your point, there's this real interesting question. When people are using this, are they assuming it's sort of like talking to another person, or are they assuming I get to play, act like talking to another person and that has real value for me. And I think that that actually is a interesting question. In therapy, right? Oftentimes in a therapist's office, you don't actually care what the therapist says. Even if they're a person. You're play acting, describing your problem to them so that you can hear yourself talk about it.
- [31:51]
Anna Wiener
Years ago, I met a man who, I think he was a graduate student who told me that when he, when he went back to school, he could no longer afford therapy. And so he, every once a week for an hour, he would sit down with a tape recorder and just talk. And I think sometimes that these products sort of serve that role for people, that it's an outlet, it feels like a safe space, judgment free. You're thinking out loud and you know, the responses can sort of buttress your own thinking in a certain way. To the, like, the question about debating with ChatGPT for hours, there's a part of me that's like, why are you debating ChatGPT for hours? You know, like, what value are you getting from that? Maybe it's to refine your own thinking, but it's, you know, like, perhaps there's no one in your life who wants to do that for a reason. But I, I don't.
- [32:38]
Tyler Foggit
The thing that I've read, it's like people basically saying like, the only way that I will change my mind about this issue is if I see a, you know, a statistic or something that suggests this. And if you're just talking to a colleague or something, they might not have those numbers on hand, but like it can do a search and then give you the piece of information that you yourself said would change your mind, which is like, I mean, that again is kind of just using it as like a high powered search engine. But like, I think there's some applications in that.
- [33:05]
Anna Wiener
Yeah, I'm making fun of it, but, you know, the colleague's like, I'll have to get back to you. But I think that these are not replacements for people. I think that a lot of the people I've spoken to are aware of that, that these are sort of supplemental. But I do think that something interesting can happen when you spend a lot of time engaging with an LLM, especially with these sort of user interfaces that are meant to be conversational and chatty and personal. And how does that affect the way that people relate to the other people in their lives, the actual people in their lives? What sort of norms does it establish or what expectations for conversation? And is there kind of like creep into your actual interpersonal relationships?
- [33:52]
Cal Newport
So, I mean, I have a response to this. I only 60% belief, but let me throw it out because it's one possibility, which is the idea that chatting with generative AI through a chat interface, that will be to the AI revolution what American online was to the Internet revolution at the time. It was like, this is what this is all about. But then you go forward 10 or 15 years and how that technology evolved is quite different. I think there's a case to be made that this is where we are with generative AI, that if you talk to people at OpenAI, for example, when they first created ChatGPT, this was a product demo, right? The model was. Their business model was API access to this thing that can understand text and produce text. And you would build programs on it, you build products on it. And they wanted to show that they were getting better at refining this. So they put out this demo. Hey, you can chat with it as like one of the types of things you could do. And then that completely caught them off guard in late 2022 when this sort of took off, right? They weren't. I happened to visit OpenAI that spring after that, and there was people running around with their hair on fire. They had no idea this was supposed to just be a demo. But it indicates, like, what they were seeing was like, what's the value here? If we have a model that can understand text, like semantically understand what you mean, it can also produce text following instructions in sophisticated ways. If you were telling me about this in, like 2015, what would you do with this technology? I would think about all the ways I would integrate into different products. I want a natural language interface into Microsoft Excel, so I don't have to figure out how to click on all these buttons to get the chart to, you know, reformat. I want to automate the processing of text that happens in the back office of our company because we have to look through all these forms. There's all these things you'd come up with, but the chat demo was very compelling. And so now that is where a lot of the space is in the room. And I think those companies don't mind it because to have a really good chatbot, you have to build huge models and there's a competitive advantage there and a mode effect and you have to have these giant data centers to do it. And it's hard to compete. But there's a possibility that 10 years from now, chatting with a bot is going to be like, oh, that's how we sort of first learned about this technology. But just like, we don't use AOL anymore. We're on apps on our phones and doing social media or whatever. It's possible. It's like now the technology is everywhere, but that has become like a less common mode. So I don't know if this is true, but it's one of these thoughts I've had. It was the thought ever since I was at OpenAI right after ChatGPT came out, this was this thought I had of like, this wasn't the plan for this technology. And so that application that could diminish in importance compared to other applications of.
- [36:17]
Charles Duhigg
This technology, like, what do you think is the. If you had to forecast, if we're having this conversation 10 years from now, is it robotics? Is it?
- [36:23]
Cal Newport
No, robotics is hard. So robotics is. I just talked to a bunch of roboticists that are like, it's a really hard problem. They're still really far away from getting any sort of meaningful generalization and model building of the world. Tesla basically canceled Optimus because it's like a very difficult problem. I thought it was going to be, we're going to see natural language interfaces more often that I can just explain what I want, as opposed to having to learn the interface that's long been the dream of computing interfaces. I think it's going to be more any place where there's textual information that needs to be processed and information pulled out or checked or forms filled out. That's a huge amount of work. This technology, you can use it, you have to make it bespoke for each of these applications, but you can use it there. So I see language model technology particular sort of as like cloud computing is just everywhere. Like, you turn it on, off the tap a little bit here, a little bit there, and it's in all sorts of products that make it more powerful as opposed to. I go to A website to interact. And when that does happen, it'll be in specialized apps. So it won't just be the ChatGPT window. But I don't know if that's true or not, but I think it's true.
- [37:31]
Tyler Foggit
We'll have more of the political scene from the New Yorker in just a moment.
- [37:42]
David Remnick
Right now, we are living through some of the most tumultuous political times our country has ever known. I'm David Remnick, and each week on the New Yorker Radio Hour, I'll try to make sense of what's happening alongside politicians and thinkers like Cory Booker, Nancy Pelosi, Liz Cheney, Tim Waltz, Ketanji Brown Jackson, Newt Gingrich, Robert F. Kennedy Jr. Charlemagne, the God, and so many more. That's all in the New Yorker Radio Hour. Wherever you listen to podcasts.
- [38:19]
Tyler Foggit
I feel.
- [38:20]
Moderator
Like at least right now, AI generated art or even like AI assisted art just has, like a real stench to it. And I think part of that is just that, like, it's been for the most part, pretty bad so far. Like, you know, you try to get it to write a New Yorker article, and it's just like, it would take me a very long time to edit that New Yorker article, basically. But I guess, by the way, it.
- [38:39]
Charles Duhigg
Takes my editor a really long time to edit my New Yorker articles too.
- [38:43]
Tyler Foggit
So that's true. Maybe in the. In the end it would be the same, but I guess, like. So I think part of that comes.
- [38:48]
Moderator
From the fact that we just, like, value human creativity and human labor, but also just the fact that the product is bad. And so I wonder, like, if we get to a point where the product is actually pretty good, do you still think that there will continue to be this stigma basically forever? Or at what point will it. Will that stigma fall away? Like, how good does the product have to be for us to be like, you know what? Actually, like, this was a pretty good AI movie. We should have more movies that are just AI.
- [39:15]
Charles Duhigg
I don't think we'll know that it's made by AI, Right. I think the artist will hide.
- [39:20]
Tyler Foggit
You have like an LLC and it's.
- [39:23]
Charles Duhigg
Yeah, I think the artist will hide. I mean, honestly, most of these Mar Vell movies, most of them are AI, right? Like, it's just like this, like, nonsense on a screen, right, that someone's using as a computer. We're aware of it, but I don't think it changes our. I don't think if someone had hand painted it instead of use pixels on screen. I don't know that people would say, like, that's much better.
- [39:43]
Anna Wiener
In defense of the films I've never seen, I think that these movies, I mean, I think actually where this conversation could go, it's. We're talking about, it's a labor conversation. I mean, the Marvel movies are products of a million tiny aesthetic judgments and decisions being made by teams of people. Just like an unbelievable number of people working on these productions. And I think that what you lose obviously when you are using an image generator or a video generator is not just the decision making process and the chain of decision making that happens in a group of people, but also obviously all of those unionized different professions. And so I don't know, I think that that's like a huge side of the story about creative industries. But Tyler, to your question about if that stigma will fall away, I think it'll take two seconds. I think that stigma will last not very long. And I say that because I think that because of the culture that we already have that the standards are not that high. And I think you'll always have people who are snobs and I support that. But, you know, that's the way it is now. And I think that we'll see a very willing audience for a lot of the stuff these generators can produce.
- [40:57]
Cal Newport
It's also worth. We don't know that this technology is going to get good enough to produce a Marvel movie from scratch. Right. So a lot of it, again, we're in that first act of Gen AI thinking is basically, if we can think of it like, because we're used to these advances, probably it'll get there. We might kind of be up against it. Right. I mean, we're thinking about the very best video generation models like OpenAI Sora 2 model, or this or that. This is a massive computational undertaking. The amount of GPUs involved is astronomical. The expense of creating each of these frames and narrative falls apart really quickly. It's very difficult to have it produce something of any length just because of the expense. There's also the scenario, if we think about fundamentally how these trained, we may never be able to produce with just a prompt a full length movie. It's also probable that writing will never get. Because we think that, oh, the writing will just get better. But if we think about how these things train, it's probably going to be difficult for a language model to produce sort of like the highest level of sort of nuanced writing with craft because again, it has to sort of win a prediction game based off of seeing huge, huge amounts of data. Most of which isn't that good. So it's the other thing I want to throw into the picture is we don't know where these technologies are going to plateau. I think we're just coming to grips with the idea that there are various plateaus. They're going to hit like GPT5 plateaued on a lot of things that we thought it was going to jump ahead. So that's the other possibility here is that we're going to reach an asymptote. So the improvements are going to at some point start to get slower and slower. This we can do better. This is helpful here. But without us having to confront it getting so good that we're like, why would we need humans? It might not catch up to us. We're not sure that that's possible.
- [42:32]
Moderator
I'm going to get to a couple audience questions before we have to end. So the first one, I mean, this is one application of AI that we didn't talk about. But how do you think about AI and political propaganda?
- [42:43]
Tyler Foggit
So I assume that some of this.
- [42:44]
Moderator
Is like, you know, the Trump Gaza video and Cuomo, his mom, Donnie.
- [42:50]
Charles Duhigg
I am actually, this is one place where I'm very pessimistic. I'm very worried that we're not going to be. Not long term. I think long term will get sophisticated. But I think in the next couple of elections you're going to see a lot of stuff. My wife loves animal videos and she sends me animal videos all the time. And I'd say 2/3 of them are created by AI. And she's so disappointed. She's a biologist, she's the person who should be like identifying this ahead of time. So I'm worried.
- [43:17]
Cal Newport
I'm very pessimistic too. The only thing that might save us here is that at least right now the technology required to produce these very good videos is extensive. So it might not be something you might need a large company to actually have the data centers large enough to produce a two minute fake political spot, which means there's potentially at least a sort of regulatory litigious response to this, where if you really tighten the screws on, if your product is like taking someone else's image and it's defamatory or this or that, you as a tech company are liable. That would be the only hope I have is that these require such large data centers and the companies are so large who can do this that at least you have some regulatory exposure that could say we're just going to have to turn these models all the Way down. If we have creative disruption and we figure out how smaller companies can build smaller models, we can do this overseas, then that would be a different story. But I think that's the only hope with the propaganda, especially deep fakes going on.
- [44:13]
Tyler Foggit
Anna, any thoughts on propaganda? And then we'll have to end it.
- [44:16]
Anna Wiener
Incredibly pessimistic. And I think it's like the worst possible time for this to be a thing. Just given media literacy and the political environment, the regulatory environment, I think it's incredibly sinister. That's the note you want to end on, though.
- [44:30]
Cal Newport
Way to bring us to a happy ending here, Tyler.
- [44:32]
Tyler Foggit
These other questions are all really depressing.
- [44:34]
Charles Duhigg
They're all downers.
- [44:35]
Tyler Foggit
Yeah, yeah, yeah. It's like, should we all get off social media?
- [44:40]
Charles Duhigg
We could talk more about animal videos.
- [44:43]
Tyler Foggit
Yeah. What's the best AI generated animal video you've seen recently?
- [44:47]
Cal Newport
Let's just play some up there.
- [44:48]
Charles Duhigg
I live in Santa Cruz, California, and there's an otter that keeps stealing people's surfboards. 841. And literally every single time I see a photo of this otter stealing a surfboard, I'm like, that has to be AI generated. That can't be true. And it's true every single time. I saw it steal a surfboard the other day. So that's the positive note I want to end on. Yeah.
- [45:09]
Jason Adam Katzenstein
All right.
- [45:10]
Tyler Foggit
Thank you so much for being here.
- [45:11]
Moderator
And thank you all for coming.
- [45:12]
Tyler Foggit
Thank you. That was Charles Duhigg, Anna Weiner, and Cal Newport. You can find their writing@newyorker.com this is the political scene from the New Yorker. I'm Tyler Foggit. Before we go today, I just want to thank you for listening to and supporting the podcast in 2025. Every Wednesday on this show, we try to make sense of American politics, even as it becomes harder to make sense of anything. And so we're very grateful to you for putting your trust in us. Thanks so much and have a very happy holidays. This episode is produced with assistance from Emily Elias and mixing by Mike Kutchman. Our producer is John Lamay. Steven Valentino is our executive producer. Our theme music is by Alison Layton Brown. We're off next week and will return on January 7th. See you in 2026.
- [46:12]
Heidi Blake
It's one of Britain's most notorious crimes, the killing of a wealthy family at White House Farm. But I got a tip that the story of this famous case might be all wrong.
- [46:25]
Cal Newport
I know there's going to be a twist one day, a massive twist at every level of the criminal justice system. There's been a cover up in this case.
- [46:32]
Heidi Blake
I'm Heidi Blake. Blood Relatives is a new series from in the Dark and the New Yorker. Find it now in the in the Dark podcast feed.
- [46:44]
Anna Wiener
From prx.