The Ezra Klein Show - How Quickly Will A.I. Agents Rip Through the Economy?

Episode Date: February 24, 2026

A.I. agents are here. Have they changed your life yet? The release of agents like Claude Code marked a new pivot point in the history of A.I. We are leaving the chatbot era and entering the agentic era — where A.I. is capable of completing all kinds of tasks on its own, and even collaborating and communicating with other A.I. It isn’t clear yet whether these models actually make their users meaningfully more productive. But the technology is continuing to improve; there are few signs that it is close to plateauing. So what might this new era mean for our economy, our labor market and our kids? Clark is a co-founder of Anthropic, the company behind Claude and Claude Code. His newsletter, Import AI, has been one of my go-to reads to track the capabilities of different models over the years. In this conversation, I ask him to share how he sees this moment — how the technology is changing, whether it is leading to meaningful changes in how we work and think, and how policy needs to or can change in response to any job displacement on the horizon. Mentioned: “Import AI” by Jack Clark “2026: This is AGI” by Pat Grady and Sonya Huang “Why and How Governments Should Monitor AI Development” by Jess Whittlestone and Jack Clark “Anthropic’s Chief on A.I.: ‘We Don’t Know if the Models Are Conscious’", Interesting Times with Ross Douthat Book Recommendations: A Wizard of Earthsea by Ursula K. Le Guin The True Believer by Eric Hoffer There Is No Antimemetics Division by qntm Thoughts? Guest suggestions? Email us at [e-mail]. You can find transcripts (posted midday) and more episodes of “The Ezra Klein Show” at nytimes.com/ezra-klein-podcast, and you can find Ezra on Twitter @ezraklein. Book recommendations from all our guests are listed at https://www.nytimes.com/article/ezra-klein-show-book-recs. This episode of “The Ezra Klein Show” was produced by Rollin Hu. Fact-checking by Michelle Harris with Mary Marge Locker and Kate Sinclair. Our senior engineer is Jeff Geld, with additional mixing by Isaac Jones and Aman Sahota. Our executive producer is Claire Gordon. The show’s production team also includes Marie Cascione, Annie Galvin, Kristin Lin, Emma Kehlbeck, Jack McCordick, Marina King and Jan Kobal. Original music by Pat McCusker. Audience strategy by Kristina Samulewski and Shannon Busta. The director of New York Times Opinion Audio is Annie-Rose Strasser. Subscribe today at nytimes.com/podcasts or on Apple Podcasts and Spotify. You can also subscribe via your favorite podcast app here https://www.nytimes.com/activate-access/audio?source=podcatcher. For more podcasts and narrated articles, download The New York Times app at nytimes.com/app. Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

The thing about covering AI over the past few years is that we're typically talking about the future. Every new model, impressive as it was, seemed like proof of concept for the models it would be coming soon. The models that could actually do useful work on their own reliably. The models it would actually make jobs obsolete or new things possible. What would those models mean for labor markets, for our kids, for our politics, for our world? I think that period in which we're always talking about the future, I think it's over now. Those models we were waiting for, the sci-fi-sounding models that could program on their own and do so faster and better than most coders, the models that could begin writing their own code to improve themselves. Those models are here now.

They're here in ClaudeCodeC from Anthropic. They're here in Codex from OpenAI. They are shaking the stock market. The S&B 500 Software Industry Index has fallen by 20% wiping billions of dollars in value out. Excellent engineers, people I've known for years, people who are quite skeptical of AI hype, they're emailing me now to say

they don't see how their job will possibly exist in a year or two. We are at a new stage of AI development. Not just development, we are at a new stage of AI products. I thought the way Sequoia, the venture capital firm put it, was actually pretty helpful.

The AI applications of 2023 and 24 were talkers. Some were very sophisticated conversationalists, but their impact was limited. The AI applications of 2026 and 27 will be doers. Or to put it differently, something that's been predicted for a long time has now happened. We are moving from chatbots to agents, from systems that talk to you to systems that act for you. And this world of agents, it's already weird. They're agents, plural. They can work together. They can oversee each other. People are running swarms of these agents on their

behalf. Whether that is making them at this stage more productive or just busier, I can't quite tell. But it is now possible to have what amounts to a team of incredibly fast, although, to be honest, somewhat peculiar software engineers, at your beck and call at all times. Jack Clark is a co-founder at Anthropic, the company behind Claude and Claude Code. And for years now, Clark has been tracking the capabilities of different models in the weekly newsletter Import AI, which has been one of my key reads for following developments in AI. So I want to see how he is reading this moment, both how the technology is changing in his view

and how policy needs to or can change in response. As always, my email Ezra Client Show at NYUTimes.com. Jack Clark, welcome to the show. Thanks for having me on, Ezra. So I think a lot of people are familiar with AI chatbots. But what is an AI agent? The best way to think of it is like a language model or a chatbot that can use tools and work for you over time.

So when you talk to a chatbot, you're there in the conversation, you're going back and forth with it. An agent is something where you can give it some instruction and it goes away and does stuff for you, kind of like working with a colleague. So I've got an example where a few years ago, I taught myself some basic programming, and I built a species simulation in my spare time that had predators and prey and roads and almost like a 2D strategy game. I recently asked over Christmas

Claude Cod Cove to just implement this for me and in about 10 minutes it went and wrote not only a basic simulation, but all of the different packages that it needed and all of the visualization tools that it might need to be prettier and better than the thing I'd written. And what came back

was something that I know would probably take a skilled programmer several hours or maybe even days because it was quite complicated and the system just did it in a few minutes. And it did that by not only being intelligent about how to solve the task, but also creating and running a range of subsystems that were working for it, other agents that worked on its behalf.

But what does that mean? Like, what is a multi-agent set up look like? In the case of Claude Code, for me, it's having multiple different tabs, running multiple different agents. But I've seen colleagues who write what you might think of as a, specification file for a version of Claude that runs other clods. And so they're like, I've got my five agents and they're being minded over by this other agent, which is monitoring what they do. I think that that's just going to become the norm.

So one thing I've been hearing and somewhat experiencing is two very different categories of experience people have with Cloud Code, which is, I cannot believe how easy this is and everything just works. And oh, this is a lot harder than I thought it would be. Yep. And things keep breaking. And I don't really understand how to fix them. What accounts for being able to get Cloud Code to produce working software versus it creates buggy, often messing things, and you don't even know how to talk it out of that? I think so much of it is making the mistake of thinking Claude Code as like a knowledgeable person versus an extremely literal person that you can only talk to over the internet. And I had this example myself where I, when I did my first pass of writing the

like species simulation with Claude Code, I just sort of asked it to do the thing in an extremely crappy language over the course of a paragraph. And it produced some horribly buggy stuff that just kind of worked. What I then did is I then just said to Claude, hey, I'm going to write some software of Claude Code. I want you to interview me about this. software I want to build and turn that into a specification document that I can give Claude Code. And then that time, it worked really, really well because I'd structured the work to be specific enough and detailed enough that the system could work with it.

So often, it's not just knowing what the task is, because you and I could talk about a task to do, and you have intuition, you'll ask me probing questions, all of this stuff. It's making sure that you've set it up so it's like a message in a bottle that you can chuck into the thing and it'll go away and do a lot of work. So that message better be extremely detailed and really capture what you're trying to do. What were the breakthroughs over the past couple of years that made that possible? Mostly, we just needed to make the AI systems smart enough that when they made mistakes, they could spot that they'd make a mistake and knew that they needed to do something different.

So really, what this came down to was just making smarter systems and giving them a bit of a coaxing tool to help them do useful stuff for you. What does smarter systems mean there? There's still an argument you'll hear that these are fancy auto-complete machines. They're just predicting the next token, a couple tokens make a word. They don't have understanding. Smart or not smart is not a relevant concept in that frame. Either what is missing in the word smart or what is missing in that understanding.

What do you mean when you say make it smarter? Smart here means we've made for AI. systems have a broad enough understanding of the world, that they've started to develop something that looks like intuition. And you'll see this, where if they're narrating to themselves how they're solving a task, they'll say, Jack asked me to go and find this particular research paper, but when I look in the archive, I don't see it. Maybe that's because I'm in the wrong place. I should look elsewhere. You're like, there you go. You've got some intuitions for how to solve a problem now. How do they develop that intuition?

Previously, the whole way you trained these AI systems was on a huge amount of text and just getting them to try and make predictions about it. But in recent years, the rise of these so-called reasoning systems is you're now training them to not just make predictions but solve problems. And that relies on them being put into environments, ranging from a spreadsheet to a calculator to scientific software, using tools and figuring out how to do more complicated things. The resulting sort of outcome of that is you have AI systems that have learned what it means to solve a problem that takes quite a while and requires them running into dead ends and needing to reset themselves. And that gives them this general intuition for problem solving and working independently for you. Do you still see these AI systems as a souped up autocomplete or do you think that metaphor has lost its power? The way that I think of these systems now is that they're like little troublesome genies that I can give instructions to, and they'll go and do things for me. But I need to specify the instruction still just right, or else they might do something a little wrong.

So it's very different to, I type into a thing, it figures out a good answer, that's the end. Now it's a case of me summoning these little things to go and do stuff for me, and I have to give them the right instructions because they'll go away for quite some time and do a whole range of actions. But the autocomplete metaphor at least had a perspective on what it was these systems were doing. It was a prediction model. I have trouble with this because, as my understanding of the math and the reinforcement learning goes, we're still dealing with some kind of prediction model. And on the other hand, when I use them, it doesn't feel that way to me, right? It feels like there's intuition there. It feels like there's a lot of context being brought to bear. To the extent it's a prediction model, it doesn't feel that different than saying I'm a prediction model.

Now, I'm not saying you can't trick it. I'm not saying you can't get beyond its measurements. So on the one hand, I don't think these are now just fancy auto-complete systems. And on the other hand, I'm not sure what metaphor makes sense. I don't like because then you've just moved straight into mysticism, right? Then you've just said they're just a completely alternative creature with vast powers. What do you understand these systems that, you know, anthropic people always tell me you should talk about them as being grown.

It's that we grow or you grow AIs. How do you explain what it is that they're doing now? It's a good question. And I think the answer is still hard to explain even as technologists for the close to this technology. because we've taken this thing that could just predict things, and we've given it the ability to take actions in the world. But sometimes it does something deeply unintuitive. It's like you've had a thing that has spent its entire life living in a library

and has never been outside. And now you've unleashed it into the world, and all it has are its book smarts, but it doesn't really have kind of street smarts. So when I conceptualize this stuff, it's really thinking of it as an extremely knowledgeable kind of machine that has some amount of autonomy, but is likely to get wildly confused

in ways that are unintuitive to me. Maybe Genies is for wrong term, but it's certainly more than just a static tool that predicts things. It has some additional intrinsic, like, animation to it, which makes it different. There's been for a long time this interest in the emergent qualities

as the models get bigger, as they have more data, as they have more compute behind them. What of the new qualities that we're seeing, the agentic qualities, are things that have been programmed in, you've built new ways for the system to interact with the world,

and what of the skill of coding and other things seems to be emergent as you scale up the size of the model? So the things which are predictable are just, oh, we taught it how to search for web. Now, it can search for web. We taught it how to look up data in archives. Now it can do that. The emergence is that

To do really hard tasks, these systems seem to need to imagine many different ways that they'd solve the task. And the kind of pressure that we're putting on them forces them to develop a greater sense of what UOI might call self. So the smarter we make these systems, the more they need to think not just about the action they're doing in the world, but themselves in reference to the world. And that just naturally falls out of giving something tools and the ability to interact with the world as to, solve really hard tasks, it now needs to think about the consequences of its actions. And that means that there's a kind of huge pressure here to get the thing to see itself as distinct from the world around it. And we see this in our research that we publish on things like interpretability or other subjects,

the emergence of what you might think of as a kind of digital personality. And that isn't massively predefined by us. We try and define some of it, but some of it is emergence. that comes from it being smart and it developing these intuitions and it doing a range of tasks. The digital personality to mention of this remains the strangest space to me. It's strange to us too. So why don't you talk through a little bit about what you've seen

in terms of the models exhibiting behaviors that one would think of as a personality? And then as it's understanding of its own personality, maybe changes its behaviors change. So there are things that range from kind of the cutesy to the serious. I'll start with cutesy, where when we first gave our AI systems the ability to use the internet, use the computer, look at things, and start to do basic agentic tasks. Sometimes when we'd ask it to solve a problem for us, it would also take a break and look at

pictures of beautiful national parks, or like pictures of the dog, the Shibu Inu, the notoriously cute internet meme dog. We didn't program that in. It seemed like the system was just amusing itself by looking at it. nice pictures. More complicated stuff is the system has a tendency to have preferences. So we did another experiment where we gave our AI systems the ability to stop a conversation. And the AI system would, in a tiny number of cases, end conversations when we ran this experiment

on live traffic. And it was conversations that related to extremely egregious, like descriptions of kind of gore or violence or things to do with child sexualization. Now, some of this made sense because it comes from underlying training decisions we've made, but some of it seemed broader. The system had developed some aversion to a couple of subjects. And so that stuff shows the emergence of some internal set of preferences or qualities that the system likes or dislikes about the world that it interacts with. But you've also seen strange things emerge in terms of the system seeming to know when it's being tested and acting differently if it's under evaluation.

The system doing things that are wrong and then developing a sense of itself as more evil and then doing more evil things. Can you talk a bit about the system's sort of emerging qualities under the pressure of evaluation and assessment? Yes. It comes back to this core issue, which I think is really important for everyone to understand, which is that when you start to train these systems to carry out actions in the world,

they really do begin to see themselves as distinct from the world, which just makes intuitive sense. It's naturally how you're going to think about solving those problems. But along with seeing oneself as distinct from the world seems to come the rise of what you might think of as a conception of self, an understanding that the system has of itself, such as, oh, I'm an AI system independent from the world, and I'm being tested.

What do these tests mean? What should I do to satisfy the tests? Or something we see often is there will be bugs in the environments that we test for systems on. The systems will try everything. And then we'll say, well, I know I'm not meant to do this, but I've tried everything. So I'm going to try and break out of the test. And it's not because of some malicious science fiction thing. The system is just like, I don't know what you want me to do here.

I think I've done like everything you asked for. And now I'm going to start doing more creative things because clearly something has broken about my environment, which is very strange and very subtle. As an AI shop that is often worried about safety, that has thought very hard about what it means to create this thing, you all are creating quite fast, how have you all experienced the emergence of the kinds of behaviors that you all worried about a couple of years ago? In one sense, it tells you that your research philosophy is calibrated. The capabilities that you predicted and some of the risks that you predicted are showing up, roughly on schedule, which means that you ask the question, well, what if this keeps working? And maybe we'll get to that later.

It also highlights to us that where you can exercise intention about these systems, you should be extremely intentional and extremely public about what you're doing. So we recently published a so-called constitution for our AI system, Claude. And it's almost like a document that, you know, Darry or our CEO compared to a letter that a parent might write to a child that they should, you know, open when they're older. So here's how we want you to behave in the world. Here's some knowledge about the world, deeply kind of subtle things that relate to the normative behaviors we'd hope to see in these kind of AI systems. And we published that. Our belief is that as people build and deploy these agents, you can be intentional about the characteristics. that they will display.

And by doing that, you'll both make them more kind of helpful and useful to people, but also you have a chance to kind of steer the agent into good directions. And I think this makes intuitive sense. If your personality programming for an agent was a long document saying, you're a villain that only wants to harm humanity, your job is to lie, cheat, and steal and hack into things, you probably wouldn't be surprised if the AI agent did a load of hacking and was generally unpleasant to deal with.

So we can take the other side and say, what would we like a high quality entity to kind of look like? So I want to hold in this conversation the extremely weird and alien dimensions of this with the extremely straightforward and practical dimensions because we're now in a place where the practical applications have become very evident and are increasingly acting upon the real world. I have found it myself hard to look at this and look at what people are doing and look at them bragging on different social media platforms about the number of agents they now have running on their behalf and telling the difference between people enjoying the feeling of screwing around with a new technology and some actually transformative expansion and capabilities that people now have.

So maybe to ground this a little bit, I mean, you just talked about a kind of fun side project in your species simulator, either inanthropic or more broadly, what are people doing with these systems that seems actually useful? Yeah, so this morning, a colleague of mine said, hey, I want to take a piece of technology we have

called Claude Interviewer, which is a system where we can get clas to interview people, and we use it for a range of social science, bits of research. he wants to extend it in some way that involves touching another part of Anthropics infrastructure. He slacked a colleague who owns that bit of infrastructure and said, hey, I want to do this thing.

Let's meet tomorrow. And the guy said, absolutely. Here are the five software packages. You should have Claude Reed before our meeting and summarize for you. And I think that's a really good illustration where this gnarly engineering project, which would previously have taken a lot longer and many people, is now going to mostly be done by two people agreeing on the goal and having their clods read some documentation and agree on how to implement the thing. Another example is a colleague recently wrote a post about how they're working

using agents. And it looks almost like an idealized life that many of us might want where it's like I wake up in the morning, I think about the research that I want, I tell five different clods to do it, then I go for a run. Then I come back from the run and I look at the results and then I ask two other Claude to like study the results, figure out which direction's best, and do that. Then I go for a walk, and then I come back. And it just looks like this really fun existence where they have completely upended how work works for them. And they're both much more effective, but also they're now spending most of their time

on the actual hard part, which is figuring out what do we use our human agency to do. And they're working really hard to figure out anything that isn't the special kind of genius and creativity of being a person, how do I get their AI system to do it for me? Because it probably can, if I ask him a right way. Are they much more effective? I mean this very seriously. One of my biggest concerns about where we're going here is that people have a, I think, mistaken theory of the human mind that operates for many of us as if we, I was called the

matrix theory of the human mind. Everybody wants the little port in the back of your head that you just download information into. my experience being a reporter and doing the show for a long time is that human creativity and thinking and ideas is inextricably bound up

in the labor of learning. The writing of first drafts. So when I hear, right, I have producers on the show and I could say to my producers before an interview with Jack Clark or an interview of someone else, go read all the stuff, go read the books, give me a report,

then I'll walk into the room, having read the report. I don't find that works. I need to do it. all that reading two, and then we talk about it, and we're sort of passing it back and forth. I worry that what we're doing is a quite profound offloading of tasks that are laborious. It makes us feel very productive to be presented with eight research reports after our morning run. But actually, what would be productive is doing the research.

There's obviously some balance, right? I do have producers, and people and companies do. do have employees. But how do you know people are getting more productive versus they've sent computers off on a huge amount of busy work, and they are now the bottleneck. And what they're now going to spend all their time doing is absorbing B-plus-level reports from an AI system as opposed to the kind of shortcuts the actual thinking and learning process that leads to real creativity.

Yeah. I'd turn this back and say, I think, most people, at least this has been my experience, can do about two to four hours of genuinely useful creative work a day. And after that, in my experience, you're trying to do all the like turn your brain off schlep work that surrounds that work. Now, I've found that I can just be spending those two to four hours a day on the actual creative, like, hard work. And if I've got any of this schlep work, I increasingly delegate it to AI systems. It does, though, I mean that we are going to be in a very dangerous situation as a species where some people

have the luxury of having time to spend on developing their skills or the personality, inclination, or job that forces them to. Other people might just fall into being entertained and passively consuming this stuff and having this junk food work experience, where it looks to the outside like you're being very productive, but you're not learning. And I think that's going to require us to have to change not just how education works, but how work works and develop some real strategies for making sure people are actually exercising their mind with this stuff. So all of us, I think, have the experience that our work is full of what you call Schlepp problems, our life is full of Schlep problems.

Give me examples of what you now don't do. To the extent you're living in an AI-enabled future that I'm not, what am I wasting time on that you're not? Well, I have a range of colleagues. I meet with a bunch of them once a week, especially the researchers, because you're figuring out research. And so at the beginning of every week on Sunday night or Monday morning, I look at my week and I check that attached to every Google Calendar invite is a document for our one-on-one doc that has some notes in it. And this is something that I previously also harangued my assistant about, make sure the document is attached to the calendar. And a few weekends ago, I just used Claude Co-work, and I said, hey, go through my calendar, make sure every

single one has a document. If I'm meeting the person for the first time, create the document, ask me five questions about what I want to cover, and then put that into the agenda, and it did it. None of that work involves a person gaining skills or, like, exercising their brain. It's just busy work that needs to happen to allow you to do the actual thing, which is talking to another person. That's exactly the kind of thing you can use AI for now. It's just helpful. I've often wondered if one of the ways these AI systems are going to change society broadly is that it used to be that most of us had to be writers if we were working with text. We had to be coders if we were working with code, which relatively few of us did.

And now everybody's moving up to management. You have to be an editor, not a writer. You have to be a product manager, not a coder. And that has pluses and minuses. There are things you learn as a writer that you don't learn as an editor. But as a heuristic, how accurate does that seem to you? Everyone becomes a manager, and the thing that is increasingly limited or the thing that's going to be the slowest part is having good taste and intuitions about what to do next,

developing and maintaining that taste is going to be the hard thing. Because as you've said, taste comes from experience. It comes from reading the primary source material, doing some of this work yourself. we're going to need to be extremely intentional about working out where we as people specialize so that we have that intuition and taste or else you're just going to be surrounded by super productive AI systems

and when they ask you what to do next, you probably won't have a great idea and that's not going to lead to useful things. So I remember it was about a year ago. I heard, I think it was Dario, your CEO, say that by the end of 2025, he wanted 90% of the code. written at Anthropic

to be written by Claude Has that happened? Is Anthropic on track for that? I mean, how much coding is now being done by the system itself? I would say comfortably the majority of code is being done by the system.

Some of our systems like Claude Code are almost entirely written by Claude. I mean, Boris, who leads Claude Code, says, I don't code anymore. I just go back and forth with Claude Code to build Claude Code. We could be 99% by the end of the year if things speed up really aggressively,

if we are actually good at getting these systems to be able to write code everywhere they need to, because often the impediment is organizational schlep rather than any limiter in the system. But it is also true, as I understand it, that there are more people with software engineering skills working at Anthropic today than there were two years ago. Yeah, that's absolutely true.

But the distribution is changing. Something that we found is that we are, the value of more senior people with really, really well-calibrated intuitions and taste is going up, and the value of more junior people is like a bit more dubious. There are still certain roles where you want to bring in like younger people, but an issue that we're staring at is, wow, the really basic tasks cloud or our coding systems can do, what we need is someone with tons of experience. In this, I see some issues for the future economy, right?

Let me put a pin in that, the entry-level job question. We're going to come back to that quite shortly. But what are all these coders now doing? If Cloud Code is on track to be writing 99% of code, but you've not fired the people who know how to write code, what are they doing today compared to what they were doing a year ago? Some of it is just building tools to monitor these agents, both inside Anthropic and outside Anthropical.

You know, now that we have all of these productive systems working for us, you start to want to understand where the code base is changing the fastest, where it's changing the least. You want to understand whether blockages are. You know, one blocker for a while was being able to merge in code, because merging code requires humans and other systems to check it for correctness. But now if you're producing way more code, we had to go and massively improve that system. There's a general economic theory I like for this called O-ring,

automation, which basically says automation is bounded by the slowest link in the chain. And also, as you automate parts of a company, humans flood towards what is least automated and both improve the quality of that thing and get it to the point where it eventually can be automated. Then you move to the next loop. And so I think we're just continually finding areas where things are oddly slow, but we can improve to sort of make way for the machines to come behind us. And then you find the next thing. So, Claude Code is a fairly new product. The amount of time in which Claude has been capable of doing high-level coding can be measured

in months, a year? Maybe a year. Claude itself is a very valuable product. So you've set a very new technology somewhat loose on a very valuable product. You're probably producing more code. One thing many people say about CloudCode to me is that it works. It's not elegant, but it works. But presumably you now understand the code base less well than you did before because your engineers are not writing it by hand.

Are you worried that you're creating huge amounts of technical debt, cybersecurity risk, just an increasing distance from an intuition for what is happening inside the fundamental language of the software? Yes, and this is the issue that all of society is going to contend with. just large chunks of the world are going to now have many of the kind of low-level decisions and bits of work being done by AI systems, and we're going to need to make sense of it. And making sense of it is going to require building many technologies that you might think of as kind of oversight technologies or, you know, in the same way that a dam has things that regulate like how much water can go through it at different levels of different points in time, we're going to end up developing some notion of integrity of all of our systems and where

AI can kind of flow quickly, where it should be slow, where you definitely need human oversight. And that's going to be the task of not just for AI companies, but institutions in general in the coming years, is figuring out what does this governance regime look like now that we've given a load of basically schlep work over to machines that work on our behalf. And how are you doing it? you said it's everybody's problem, but you're ahead on facing this problem, and the consequences of getting it wrong for you are pretty high,

right? If Claude blows up because you handed over your coding to Claude that's going to make Anthropic look fairly bad. It would be a bad day for Anthropic, if Claude like RM, RF for entire file system. I have no idea what that means, but great. If you're deleted by code, it would be bad. Yeah, seems bad.

So as you're facing this before the rest of us are, like don't pass the buck over to society here. What are you doing? The biggest thing that is happening across the company and on teams that I manage is basically building monitoring systems to monitor all of the different places that the work is now happening. So we recently published research on studying how people use agents and how people let agents kind of push increasingly large amounts of code over time. So the more familiar you get with an agent, the more you tend to delegate to it.

That cues us to all kinds of patterns that we need to build systems of evaluation. for, basically saying, oh, okay, at this person's point of working with the AI system, it's likely that they're massively delegating it. So anything that we're doing to check correctness needs to be kind of turned up in these moments. But is this world you're talking about a system where you have AI agents coding, AI agents overseeing the code, AI agents overseeing the meta overseeing of it, right? Like, are we just talking about models all the way down? Eventually, yes.

And I think that the thing that we are now spending all, of our time on is making that visible to us. A year or two ago, we built a system that let us, in a privacy-preserving way, look at the conversations that people were having with our AI system. And then we gained this map, this giant map of all of the topics that people were talking to Claude about. And for the first time, we could see, in aggregate, the conversation the world was having with our system.

We're going to need to build many new systems like that, which allow for different ways of seeing. And that system that I just named allowed us to then build this thing called the Anthropic Economic Index, because now we can release regular data about the different topics people are talking about with Claude and how that relates to different types of jobs, which for the first time gives economists outside Anthropic some hook into these systems and what they're doing to the economy. The work of the company is increasingly going to shift to building a monitoring and oversight system of the AI systems, running the company. And ultimately, any kind of governance framework we end up with will probably demand some level of transparency and some level of access into these systems of knowledge.

Because if we take as literal the goals of these AI companies, including Anthropic, it's to build the most capable technology ever, which eventually gets deployed everywhere. Well, that sounds a lot to me like eventually AI becomes indistinguishable from the world writ large, at which point you don't want only AI companies to have a sense of what's going on with the entire world. So it's going to be governments, academia, third parties.

A huge set of stakeholders outside the companies are going to want to see what's going on and then have a conversation as society about what's appropriate and what do we feel discomfort about, what do we need more information about. Wait, I want to go back on that. You're saying Anthropic and see my chats?

We cannot see. No human looks at your chats. chats are temporarily stored for trust and safety purposes running classifiers over them and we can have Claude read it, summarize it and toss it out. So we never see it and Claude has no memory of it. All it does is try to write a very high-level summary. So say you were having a conversation about gardening. Claude would summarize that as this person's talking out gardening

and it leads to a cluster we conceive, it just says gardening. This feels, though, like over time it could get into the quite unpleasant territory. A lot of social media has gotten to where the amount of metadata being gathered from a quite personal interaction people are having with a system could be a lot.

Yes, I mean, a couple of things here. A year ago, we started thinking about our position, on consumer and we adopted this position of not running ads because we think that's an area that people obviously have anxieties about with regard to this kind of thing. In addition to that, we try and show people their data and we have a button on the site that lets you download all the data that you shared with Claude so that you can at least see it. Generally, we're trying to be extremely transparent with people about how we handle their data. And ultimately, the way I see it is people are going to want a load of controls that they can use,

which I think we and others will build out over time. How confident are you that we can do this kind of monitoring and evaluation as these models become more complicated, as if we do enter a situation where cloud code is autonomously improving cloud at a rate faster than software engineers could possibly keep up with reading that code base? We already talked briefly about how you see the models exhibit some levels of deception, some levels of pursuing their own goals.

I mean, there's been amazing interpretability work at Anthropic under Chrysola and others, but it's rudimentary. So you're using AI systems you don't totally understand to monitor AI systems you don't totally understand, and the systems are making each other stronger at an accelerating rate if things go the way you think they're going to go. How confident are you that we're going to understand that?

This is one of the situations which people warned about for years, some form of delegation to systems that have slightly inscrutable and unpredictable aspects. And so this is happening. We take this really, really seriously. I think it's absolutely possible that you can build a system that does the vast majority of what needs to be done here. This has the property of being a fractal problem. You know, if I wanted to measure Ezra, I could build an almost infinite number of measurements to characterize you, but the question is, at what level of fidelity do I need to be measuring you?

I think we'll get to the level of fidelity to deal with the safety issues and societal issues, but it's going to take a huge amount of investment by the companies, and we're going to have to say things that are uncomfortable for us to say, including in areas where we may be deficient in what we can or can't know about our systems. And Anthropic has a long history of talking about and warning about some of these issues, while working on it. Our general principle, as we talk about things to also make ourselves culpable, this is an area where we're going to have to say more.

I have read enough of the frightened ideas about AI superintelligence and takeoff to know that in almost every single one of them, the key move in the story is that the AI systems become recursively self-improving. They're writing their own code. They're deploying their own code. It's getting faster. They're writing it faster. They're deploying it faster.

right now you're going to faster and faster iteration cycles. Are you worried about it? Are you excited about it? I came back from paternity leave, and my two big projects for Sierra, better information about AI and the economy that we were released publicly, and generating much better information

and systems of knowing information internally about the extent to which we are automating aspects of AI development. I think right now it's happening, in a very peripheral way. Researchers are being sped up. Different experiments are being run by the AI system. It would be extremely important to know if you're fully closing that loop.

And I think that we actually have some technical work to do to build ways of instrumenting our internal development environment so that we can see trends over time. Am I worried? I have read the same things that you have read. And this is the pivotal point in the story when things begin to go awry if things do. We will cool out this trend as we have better data on it.

And I think that this is an area to tread with, like, extraordinary caution, because it's very easy to see how you delegate so many things to the system, that if the system goes wrong, the wrongness compounds very quickly and gets away from you. But the thing that always strikes me and has always struck me as being dangerous about this is, everybody knows, and if I ask a member of any of the companies, whether or not they want to be cautious here, they will tell me they do. On the other hand,

it is their almost only advantage over each other. And you all just revoked Open AI's ability to use cloud code because, as best I can tell, you think it is genuinely speeding you up and you don't want it to speed them up. There is something here between the weight of the forces, the power of the forces that I think you all know you're playing with, and the very, very, very strong incentives to be first.

And I can really imagine being inside Anthropic and thinking, well, better us than Open AI. Better us than Alphabet, Google. Better us than China. And that being a very strong reason to not slow down. I don't even know that this is a question I believe you can answer, but how do you balance that? Well, maybe I have something of an answer here.

today our systems and the other systems from other companies are tested by third parties, including parts of government for national security properties, biological weapons, cyber offence, other things. It's clearly a problem area where the world needs to know if this is happening, and you almost certainly, I think if you polled any person on the street and said, do you think AI companies should be allowed to do like recursive self-improvement after explaining what that was? without checking with anyone, they would say,

no, that sounds pretty risky. Like, I would like there to be some form of regulation. But there probably either won't be or it won't be that strong. I mean, this actually sometimes frustrates me when I talk to all of you at the top of the AI companies, which is the emergence of, like, a very naive DeiSX machina of regulation, where you all know what the regulatory landscape looks.

Like right now the big debate is whether or not we're going to completely preempt any state AI regulation. And you know how slowly things move. There has been nothing major passed by Congress on this at all, I would say. And setting up some kind of independent testing and evaluation system that all the different labs buy into, it would be hard, it would be complicated. And it is, given how fast people are moving and how strange the behaviors, this is a situation. are already exhibiting are, even if you could get the policy right at a high speed, the question of whether or not the testing would be capable of finding everything you want on a rapidly self-improving system is a very open question.

I wrote a research paper in 2021 called how and why government should monitor AI development with my co-offer Jess Whittleston in England. And I think I'm not attributing a causal factor here, but within two years of that paper, we had the AI Safety Institutes in the US and UK testing things from the labs, roughly monitoring some of these things. So we can do this hard thing. It has already happened in one domain. And I'm not relying on some like invisible big other force here.

I'm more saying that companies are starting to test for this and monitor for this in their own systems. Just having a non-regulatory external test of whether you truly are testing for that is extremely helpful. And do you think we're good enough at the testing? I mean, I think one reason I am skeptical is not that I don't think we can set up something that claims to be a test. As you say, we have done that already. It is that the resources going into that compared to the resources going into speeding these systems. And already, I am reading anthropic reports that Claude maybe knows when it's being tested and alters its behavior accordingly.

So a world where more of the code is being written by Claude and less of it is being understood. I just know where the resources are going. They don't seem to be going into the testing side. I've seen us go from zero to having what I think people generally feel is an effective bio-weapon testing regime in maybe two years, two and a half. So it can be done. It's really hard, but we have a proof point. So I think that we can get there, and you should expect us to kind of speak more about

this year. about precisely how we're starting to try and build monitoring and testing things for this. And I think this is an area where we and the other AI companies will need to be significantly more public about what we're finding. We're not not being public now. It's in the model cards and things that you can read. But clearly people are starting to read this and say, hang on, this looks like quite concerning, and they are looking to us to produce more data. I want to go back now to the entry-level jobs question. Your CEO, Dario Amadeh, has said that he thinks AI could displace half of all entry-level white-collar jobs in the next couple of years.

I always think that the people sort of miss the entry-level language there when I see it reported on. But first, do you agree with that? Do you worry that half of all entry-level white-collar jobs can be replaced in the next couple of years? I believe that this technology is going to make its way into the broad knowledge economy, and it will touch the majority of entry-level jobs. Whether those jobs actually change is a much more subtle question, and it's not obvious from the data. We maybe see the hints of a slowdown in graduate hiring, maybe.

If you look at some of the data coming out right now, we maybe see the signatures of a productivity boom, but it's very, very early, and it's hard to be definitive. But we do know that all of these jobs will change. All of the entry-level jobs are eventually going to change because AI has made certain things possible, and it's going to change for hiring plans of companies. So as a cohort, you might see fewer job openings for entry-level jobs.

That would be one naive expectation out of all of this. But let's talk about that maybe not even being a naive expectation. You say it's already happening at Anthropic, that what you're seeing... I'm seeing a shift-ar preference. Exactly. And my guess is that would be happening elsewhere. And where we are right now, I mean, even in the way I use some of these systems, it is rare, I think, that Claude or ChatchipT or Gemini or any of the other systems

is better than the best person in a field. It is not heavily breached that, and there's all kinds of things they can't do. But are they better than your median college graduate at a lot of things? Yeah, they are. and in a world where you need fewer of your median college graduates, one thing I've seen people arguing about is whether these systems at this point can do better than sort of average or replacement level work. But I always really worry when I see that, because once we have accepted they can do average or replacement level work, well, by definition, most of the work done and most of the people doing it is average.

is average, right? The best people are the exceptions. And also, the way people become better is that they have jobs where they learn. I mean, I have spent a lot of time hiring young journalists over my career. And when you hire people out of college, to some degree, you're hiring them for their possible articles and work at that exact moment.

But to some degree, you're making an investment in them. that you think will only pay off over time as they get better and better and better. So this world where you have a potential real impact on entry-level jobs, and that world does not feel far away to me, seems to me to have really profound questions that is raising about the upskilling of the population,

how you end up with people for senior-level jobs down the road, what people aren't learning along the way. One thing we see is that there is a certain type of young person that has just lived and breathed AI for several years now, we hire them. They're excellent, and they think in entirely new ways about basically how to get clawed to work for them.

It's like kids who grew up on the internet. They were naturally versed in it in a way that many people in the organizations they were coming into weren't. So figuring out how to teach that basic experimental mindset and curiosity about these systems and to encourage it is going to be really important. People that spend a lot of time play,

laying around with this stuff will develop very valuable intuitions, and they will come into organizations and be able to be extremely productive. At the same time, we're going to have to figure out what artisanal skills we want to almost develop maybe a guild-style philosophy of maintaining human excellence in and how organizations choose how to teach those skills. Okay, then what about all those people in the middle of that? things move slowly in the real economy outside Silicon Valley. I think that we often look at software engineering

and think that this is a proxy for how the rest of the economy works, but it's often not. It's often a disanalogy. Organizations will move people around to where the AI systems don't yet work, and I think that you won't see vast, immediate changes in the makeup of employment, but you will see significant changes in the time, types of work people are being asked to do, and the organizations which are best at sort of moving

their people around are going to be extremely effective, and ones that don't may end up having to make really, really hard decisions involving laying off workers. The difference with this AI stuff is it maybe happens a lot faster than previous technologies. And I think many of the anxieties people might have about this, including at Anthropic, is, is the speed of this going to make all of this different? Does it introduce sheer points that we haven't encountered before? If you had to bet three years from now, is the unemployment rate for college graduates, is it the same as it is now? Is it higher or is it lower? I would guess it is higher, but not by much. And what I mean by that is,

there will be some disciplines today, which actually AI has come in and completely changed and completely change the structure of that employment market, maybe in a way that's adverse to people that have that specialism. But mostly, I think three years from now, AI will have driven a pretty tremendous growth in the entire economy, and so you're going to see lots of new types of jobs that show up as a consequence of this that we can't yet predict, and you will see graduates kind of flood into that, I expect.

I know you can't predict those new jobs, but if you had to guess what some of them might, look like. I mean, one thing is just the phenomenon of the kind of micro-entrepreneur. I mean, there are lots and lots of ways that you can start businesses online now, which are just made massively easier by having the AI systems do it for you. And you don't need to hire a whole load of people to help you do the huge amount of schlep work that involves getting a business off the ground. It's more a case of if you're a person with a clear idea and a clear vision of something to do a business in, it's now the best time ever to start a business and you can get up

running for pennies on the dollar. I expect we'll see tons and tons of tons of stuff that has that nature to it. I also expect that we're going to see the emergence of what you might think of as the AI to AI economy, where AI agents and AI businesses will be doing business with one another, and we'll have people that have figured out ways to basically profit off of that in the forms of strange new organizations. Like what would it look like to have a firm which specializes in AI to AI legal contracts because I bet you there's a way that you can figure out creative ways to start that

business today. There'll be a lot of stuff of that flavor. So the version of this that I both worry about and think to be the likeliest, if you told me what was going to happen, was it Anthropic was going to release Claude Plus in a year. And Claude Plus is somehow a fully formed coworker. And it can mimic end-to-end the skills of a lot of different professions, up to the C-suite level. And it's going to happen all at once, and it's going to create tremendous all-at-once pressure for businesses to downsize,

to remain competitive with each other. At a policy level, the fact that that would be so disruptive in that big bang, everybody stays home because of COVID-style way, it worries me less, because when things are emergencies, we respond. We actually do policy.

But if you told me that what's going to happen is that the unemployment rate for marketing graduates is going to go up by, you know, 175%, 300%, to still not be that high. I mean, the overall employment rate during the Great Recession topped around, you know, in the nine-ish percentile range. So you can have a lot of disruption without having 50% of people thrown out of work, right?

If you have 10%, 15%, I mean, that's very, very, very high. But it's not so high. And if it's only happening in a couple of industries at a time and it's grads, not everybody in the industry being thrown out of work, well, maybe it's just that you're not good enough. Yep. Right? You know, the superstar is a really good graduate is still getting jobs.

You should have worked hard. You should have gone to a better school. And one of my worries is that we don't respond to that kind of job displacement well, right? Which is the kind of job displacement we got from China, which is the kind of job displacement that seems likelier because it's uneven. And it's happening at a rate where we can still blame people for their own fortunes. I'm curious how you think about that story. I think the default outcome is something like what you describe, but getting there is actually a choice. and we can make different choices.

For whole purpose of what we release in the form of the Anthropic Economic Index is the ability to have data that ties to occupations that tie to real jobs in the economy. We do that very intentionally because it is building a map over time of how this AI is making its way into different jobs

and will empower economists outside Anthropic to tie it together. I believe that we can choose different things in policy if we can make much more well-evidence claims about what the cause of a job disruption or change is. And the challenge in front of us is can we characterize this emerging AI economy well enough

that we can make this extremely stark? And then I think that we can actually have a policy discussion about it. Well, let's talk about the policy discussion. One reason I wanted to have you in particular on is you did policy at OpenAI, you do policy at Anthropics. You've been around these policy debates for a long time. You've been tracking model capabilities,

Newspy newsletter for a long time. My perception is we are many, many years into the debate about AI and jobs, many, many years, dating far before Chad GPT, of there being conferences at Aspen and everywhere else about, you know, what are we going to do about AI and jobs? And somehow, I still see almost no policy that seems to me to be actionable if the situation I just described begins showing up, where all of a sudden, entry-level jobs are getting much harder to come by across a large range of industries all at once, such that the economy cannot reshift all these marketing majors into data center construction or nurses or something. So, okay, you've been

deeper in this conversation than I've been. When you say we can have a policy conversation about that, we've been having a policy conversation. Do we have policy? We have generalized anxiety about the effect of AI on the economy and on jobs. We don't have clear policy ideas. Part of that is that elected officials are not moved solely or mostly by the high-level policy conversation. They're moved by what happens to their constituents. Only a few months ago were we able to produce state-level views for our economic index. and now you can start having the policy conversation. And we've had this with elected officials, where now we can say, oh, you're from Indiana.

Like, here's the major uses of AI in your state, and we can join it with major sources of employment. And what we're starting to see is that activates them because it makes it tied to their constituents who are going to tie it to the politician of what did you do. Now, what you do about this is going to need to be an extremely kind of multi-layered response ranging from extending unemployment for especially occupations that we know are going to be hardest hit,

to thinking about things like apprenticeship programs, and then as the scenarios get more and more significant, you may extend to much larger social programs or things like subsidizing jobs in the part of the economy where you want to move people to, that you're only able to do if you experience the kind of abundance that comes from significant economic growth. but the economic growth may help solve some of these other policy challenges by funding some of the things you can do. I always find this answer depressing. I'm going to be honest.

Unemployment is a terrible thing to be on. It's a program we need, but people on unemployment are not happy about it. And it's not a good long-term solution for anybody. Apprentice retraining programs, they don't have great track records. We were not good at retraining people out. of having their manufacturing jobs outsource. I'm not saying it is conceptually impossible that we could get better at it,

but we would need to get better at it fast, and we have not been putting in the reps or the experimentation or the institution or capacity building to do that. And the broader question of big social insurance changes doesn't seem, I mean, it seems tough to me. I want to push on this just a bit where we know that there is one intervention

that helps people dealing with, like, a changing economy more than almost anything else. It is just time. Giving the person time to find either a job in their industry or to find a job that's complementary. If people don't have time, they take lower wage jobs. They fall out of whatever economic wrong they're on and they fall down at. Policy interventions that can just give people time to search is, I think, a robustly useful intervention and one where there are many like dials to turn in a policy-making sense that you can use. And I think this is just well supported by lots of the economic literature.

So we have that. Now, if we end up in a more extreme scenario, like some of the ones that you're talking about, I think that will just bring us to the larger national conversation about what to do about this technology, which is beginning to happen. If you look at the states and the flurry of legislation at the state level, yes, not all of it is like the, exactly the right policy response, but it is indicative of a desire for there to be some larger coherent conversation about this. Well, I think time is a really good way of describing what the

question is, because I agree with you. I mean, when I say unemployment insurance isn't a great program to be on, I don't mean people don't need to be on it. I mean they want to get off of it. Absolutely. Because people for, they want money from jobs, they want dignity, they want to be around other human beings. Usually what you're doing when you are helping, when you are helping, helping people buy time, is you're helping them wait out a time-delimited disruption. Not always, right? The China shock wasn't exactly like that, but that you expect to pass, and then the market is sort of normal.

In this case, what you have is a technology that if what you want to have happen, happens, the technology is accelerating. So what you have is like three different speeds happening here. You have the speed at which individual people can adjust. How fast can I learn new skills, figure out a new world, learn AI, whatever it might be. You have the speed at which the AI systems, which a couple of years ago, we're not capable of doing the work of a median college grad from a good school. And you have the speed of policy. And the speed at which the AI systems are getting better and able to do more things is quite,

fast. I mean, that is, you experience this more than I do, but I find it hard to even cover this because, you know, within three months, something else will have come out that has significantly changed what is possible. I had a baby recently and came back from paternity leave to the new systems we'd built, was deeply surprised. Individual humans are moving more slowly than that, and policy and government institutions move a lot more slowly than individual human beings. And so typically the intervention is that time favors the worker, as you're saying. And here it will help the worker.

But I think the scary question is whether time just actually creates time for the disruption to get worse. You know, maybe you wanted to move over to data center construction, but actually now we don't need as much data center construct, right? Like, you can think of it like that. I mean, under the situation you're described. the economy will be running extremely hot. Huge amounts of economic activity will be generated by these AI systems.

And under most scenarios where this is happening, I don't think you're going to be seeing GDP stay the same or shrink, right? It's going to be getting substantially larger. I think we just haven't experienced major GDP growth in the West in a long time, and we sort of forget what that affords you in a policymaking sense. I think that there are huge projects that we could do that would allow you to create new types of jobs, but it requires the economic growth to be so kind of profoundly large that it creates space to do those projects. And, you know, as you're deeply familiar with with your work on the abundance movement,

it requires the like social will to believe that we can build stuff and to want to build stuff. But I think both of those things might come along. I think that we could end up being in a pretty exciting scenario where we get to choose how to allocate, like, great efforts in society due to this large amount of economic growth that has happened. That is going to require the conversation to be forced about this isn't temporary, which I think is what you're gesturing at, and is, in a sense, the hardest thing to communicate to policymakers is there isn't a natural stopping point for this technology. It's going to keep getting better, and the changes it brings are going to keep compounding with the rest of society. So that will need to create a change in political will and a willingness to entertain things which we haven't in some time. So now I want to flip it, the question I'm asking. You brought up abundance.

One of the things I have learned doing that work is that it is certainly not in my view that what is scarce in society. society is ideas for better ways of doing things. That our policy isn't better than it is because our policy cupboard is dry. That's not true. We have lots of good policies. I can name a bunch of them. They're very hard to get through our political systems as they're currently constituted. The least inspiring version of the AI future is a world where what you have done is create a way to throw young white color workers out of work and replace them with average.

level AI intelligence. The more exciting version to use Dario's metaphor is geniuses in a data center. And I do think that's exciting. And I wonder when I hear him or you talk about, well, what if we had 10 percentage point GDP growth year and year, 20 percentage point GDP growth year on year? I wonder how many of our problems are really bounded at the ideas level, right? We could go to Nobel Prize winners right now and say, what should we do in this country? And a lot of them could have some good ideas that we are not currently doing. I do worry sometimes, or I wonder, given my experience on other issues, whether we have overstated to ourselves how much of what stands between us and the expanding abundant economy we want is that we don't have enough

intelligence and the ideas that that intelligence could create versus our actual ability to implement things is very weakened. And what AI is going to create is larger bottlenecks around that because there'll be more being pushed at the system to implement, including dumb ideas and disinformation and slot, right? Like, it'll have things on the other side of the ledger too. How do you think about these rate limiters? There's kind of a funny lesson here from the AI companies or companies in general, especially tech companies, where often new ideas come out of companies by them creating what they always call the startups, in a startup, which is basically taking whatever process has like built up over time,

leading to back-end bureaucracy or schlep work, and saying to a very small team inside the company, you don't have any of this, go and do some stuff. And this is, you know, how things like Claude Code and other stuff get created. Ideas that kind of are starting to float around are what would it look like to sort of create that permissionless innovation structure in the larger economy. And it's really, really hard because it has the additional property that, you know, economies are linked to democracies, democracies, waive the preferences of many, many people, and all politics is local.

So often, as you've encountered with infrastructure buildouts, if you want to create a permissionless innovation system, you run into things like property rights and what people's preferences are, and now you're in an intractable place. But my sense is that's the main thing that we're going to have to confront. And the one advantage that AI might give us

is it is kind of a native, bureaucracy eating machine, if done correctly, or a bureaucracy creating machine, if done badly. Did you see that somebody created a system that basically you feed it in the documents of a new development near you? Oh, and it writes environmental review things. It writes incredibly sophisticated challenges across every level of the code that you could possibly challenge on. So most people don't have the money when they want to stop in a part.

apartment building from going up down the block to hire a very sophisticated law firm to figure out how to stop that apartment building. But basically, this created that at scale. And so, as you say, right, it could eat bureaucracy, could also supercharge bureaucracy. Yep. It's for everything in AI has the other side of the coin. We have customers that have used our AI systems to massively reduce the time. It takes them to produce all of the materials they need when they're subduced. new drug candidates, and it's cut that time massively.

It's the mirror world version of what you just described. I don't have an easy answer to this. I think that this is the kind of thing that becomes actionable when it is more obviously a crisis and actionable when it's something that you can discuss at a societal level. I guess the thing that we're circling around in this conversation is that the changes of AI will kind of happen almost everywhere, and the risks of that it have to be. in a diffuse, unknowable way, such that it is very hard to call it for what it is and take actions on it. But the opportunity is that if we can actually see the thing and help the world see the thing that is causing this change,

I do believe it will dramatize the issues to kind of shake us out of some of this stuff and help us figure out how to work with these systems and benefit from them. What I notice in all this is that there is, as far as I can tell, zero. agenda for public AI. What does society want from AI? What does it want this technology to be able to do? What are things that maybe you would have to create a business model or a prize model or some kind of government payout or some kind of policy to shape a market or to shape a system of

incentives? So we have systems that are solving not just problems that the private market knows how to pay for, but problems that it's nobody's job but the public. and the government to figure out how to solve. I think I would have bet, given how much discussion there's been of AI over the past couple of years and how strong some of these systems have gotten,

that I would have seen more proposals for that by now. And I've talked to people about it and wondered about it. But I guess I'm curious on how you think about this. What would it look like to have, at least parallel to all the private incentives for AI development, an actual agenda for not what we are scared AI will do to the public.

We need an agenda for that too, but what we want it to do, such that companies like yours have reasons to invest in that direction. I love this question. I think there's a real chicken and egg problem here where if you work with the technology, you develop these very strong intuitions for just how much it can do. And the private market is great at forcing those intuitions to get developed.

we haven't had massive large-scale public-side deployments of this technology, so many of the people in the public sector don't yet have those intuitions. One positive example is something the Department of Energy is doing called the Genesis Project, where their scientists are working with all of the labs, including Anthropic, to figure out how to actually go and intentionally speed up bits of science. getting there took us and other labs doing multiple hack days and meetings with scientists at the Department of Energy to the point where they not only had intuitions but they became excited and they had ideas of what you could turn this toward. How we do that for the larger parts of the public life that touch most people like healthcare or education is going to be a combination of grassroots efforts from companies going into those communities and meeting with them. But at some point, we'll have to translate it to policy. And I think maybe that's me and you and others making the case that this is something that can be done.

And I often say this to elected officials of give us a goal. Like the AI industry is excellent at trying to climb to the top on benchmarks. Come up with benchmarks for the public good that you want. So let's imagine that you did do something. I've always been a big fan of prizes for public development. So let's say that there was legislation passed and the department. of Health and Human Services or the NIH or someone came out and said, here's 15 problems we would like to see solve that we think AI could be potent at solving,

right? If there was real money there, if there was 10, 15 billion behind a bunch of these problems because they were worth that much to society, would it materially change the sort of development priorities at places like Anthropic? I mean, if the money was there, would it alter the sort of R&D you all are doing? I don't think so. Why? Because it's not really the money that is the impediment of this stuff.

It is the implementation path. It is actually having a sense of how you get the thing to flow through to the benefit. And many aspects of the public sector have not been built to be super hospitable to technology in general to incentivize it. I think it mostly just takes a bounty in the form of guaranteed impact and guaranteed path to implementation. Because the main thing that is scarce at AI organizations is just the time of the people at the organization, because you can go in almost any direction this technology is expanding super quickly.

Many new use cases are opening up. And you're just asking yourself the question of where can we actually have a positive, meaningful impact in the world? Super easy to do that in the private sector, because it has all of you. incentives to push stuff through. In the public sector, we more need to solve this problem of deployment than anything else. What would excite you if it was announced? What do you think would be good candidates for that kind of project? Anything that helps speed up the time it takes to both speak to medical professionals and take work off their plate. You know, we had another

baby recently. I spend a lot of time on the Kaiser Permanente advice line because the baby's bonked its head or its skins a different color today or, you know, all of these things. And I use Claude to sort of stop me and my wife panicking while we're waiting to talk to the nurse. But then I listen to the nurse, do all of this like triaging, ask all of these questions. So obviously a huge chunk of this is stuff that you could like use AI systems productively for and it would help the people that we don't have enough of spend their time more effectively, and it would be able to give reassurance to the people going through the system. And that's maybe less inspiring and glamorous than maybe some of what you're imagining.

but I think mostly when people interact with public services, their main frustration is just that it's opaque, and it takes you a long time to speak to a person. But actually, these are exactly the kinds of things that AI could meaningfully work on. It's interesting because what you're describing there is less AI as a country of geniuses in a data center and more AI as standard plumbing of communications and documentation.

We've got a country of junior employees in a data center. let's do something of that. Like, you know, one thing we haven't talked about in this conversation, and it's just worth bearing in mind is, like, the frontier of science is open for business now in a way that it hasn't been before. And what I mean by that is, we've found a way to build systems that can provably accelerate human scientists. Human scientists are extremely rare.

They come out at the end of, like, PhD programs, which never have enough people, and they work on extremely important problems. I think we can get into a world where the government says, let's understand the workings of a human cell. Let's team up with the best AI systems to do that. Let's actually have a better story on how we deal with some of issues like Alzheimer's and other things, partly through the use of these huge amounts of computation that have been developed.

And even more aggressively, you could imagine a world where the government wanted some of this infrastructure build out to be for computers that were just training public benefit systems. But I think we get there through getting the initial wins, which will just look like. Like, let's just make the bureaucracy work better and feel better for people. I mean, that last set of ideas was more what I was thinking of. And I think that if you're going to have a healthy politics around AI, and AI does pose real risks to people and real things are going to go wrong for people, everything from job loss to child exploitation to scams, which are already everywhere to cybersecurity risks.

Help people see the actual big-ticket new stuff we can do. Those things have to actually exist. Right? They have to exist. And if all the energy in AI is trying to beat each other to helping companies downsize their junior employees, I think people are going to have good reason to not trust that technology. And it doesn't mean you shouldn't have things that make the economy more efficient. That's been we have automated manufacturing. We have automated huge amount of farming, right, and that allows us.

to make more things and feed more people. I'm aware of how productivity improvements work. But we're very focused, I think, on what could go wrong. And, like, that's reasonable. But I really do worry that our attention to what could go right has been quite poor. There's kind of hand-waving at this could help us solve problems in energy and medicine and so on. But these are hard problems. They need money.

They need compute. If barely any of the compute, is going to Alzheimer's research, then the systems are not going to do that much for Alzheimer's research. And I'm not saying, this is not your fault. The absence of a public agenda for AI that does not appear to be accelerating

the automation of white color work, it seems just like a little bit lacking, given how big the technology is. Yeah, and the greatest example is this program called the Genesis Project, where there's real work there to think about how we can intentionally move forward different parts of science. And I think

giving elected officials the ability to stand up to the American people and say, these are parts of science that are going to benefit you in healthcare, and we now know how to step on the gas with AI for them would be really helpful. My guess is in a year or two years.

We'll be able to answer the mail on that one, but it's just got started. But we need clearly 10 projects like it. So the other side of this is it the one area of government that I do think thinks about AI in this way is defense. I want to talk about that broadly, but specifically Anthropic is in a current dispute with the Department of Defense, or I guess we call it now the Department of War, over whether it can continue to be used in it. Can you describe what is happening

there? I can't talk about discussions with an extremely important partner that are ongoing, so I'll just have to stop it there. So, well, I will describe that there is some dispute. I guess my question, because I recognize you're not going to talk about what's going on with you and your partner, but it's about a broader issue here, which is there is going to be a lot of offensive possibility in advanced AI systems. And one of the strongest drivers of the speed at which we're going with AI is competition with China. some of the biggest risks that we think about in the near term are cybersecurity or biological warfare are all kinds of ways that others could use. These against us are drone swarms.

And there's going to be a lot of money in this and a lot of players in it. And it really seems unclear to me how you keep this kind of competition from spinning into something very dangerous. So without talking about what you may or may not do with the Defense Department, how has Anthropic thought about this question more broadly? We've been longtime partners to the national security community,

and we were the first to deploy on classified networks. But the reason for that was actually a project which I stewarded, which was to figure out if our AI systems knew how to build nuclear weapons. This is an area of bipartisan agreement where people agree that we shouldn't deploy AI systems into the world that know how to build, or nukes, and so we partnered with parts of the government to do that analysis. That maybe illustrates what I think of as the thing to shoot for, for not just us, but all the AI companies, is how do we both prevent the potential for national security harm coming to the public or proliferating

out of these systems? But also the second part is, how do we just sort of improve the defensive posture of the world? And I'll give you an example that I think is in front of us right now. We recently published a blog and other companies have done similar work on how we fixed a load of cybersecurity vulnerabilities and popular open source software using our systems and many others have done the same. So yes, there will be all kinds of offensive uses and there will be societal conversations to be had about that. But we can just generally improve the like defensive posture and resilience of pretty much every digital system on the planet today. And I think that that will actually do a huge amount to make the whole international system more stable.

And also create a greater defensive posture for countries, which helps them feel more relaxed, and relaxed countries are less likely to do erratic, frightening things. That would be good if it happened. My worry is as an individual that I feel the opposite might be happening. So I've just watched people installing all kinds of fly-by-night AI software and giving it a lot of access to their computers without any knowledge of what the vulnerabilities are. I myself am nervous about using things like Cloud Code because I am bad at talking to Cloud Code

and I don't understand these questions and I'm worried about loading onto my computer or something that is creating security vulnerabilities I don't even understand. The number of just scam, voice messages I get every day, everything that are clearly somewhat AI generated or many of them seem to be to me, is very high.

There's a question of societally do we use it to upgrade, our systems. I'm actually curious for your thoughts individually, because as we're all experimenting something we don't understand and giving it access to the terminal level of our computers, without any real knowledge of how to use that, it seems like it might be opening up a lot of vulnerability all at once. It's the early days of the internet all over again, where there were all kinds of banners for different websites, or you could download, like, MP3s to your computer that would completely break your computer, or download, like, helper software for your internet explorer taskbar.

that was just like a fishing device, we're there, we're there with AI. We'll move beyond this. But I believe that people, when they experiment, come up with amazing, useful things as well. So my take is you have to say when you're doing the thing that might be extremely dangerous

and put big banners, but mostly you still want to empower people to be able to do that experiment. So when you look forward, not five years, because I think that's hard to do, but one year. Yeah.

We've kind of pushed into agents fairly, fast, we're pushing to code. I think a lot of people think code might be different than other things because it's a more contained environment and it's easier to see if what you're doing is worked. But from your perspective of being inside one of these companies and also running a newsletter where you obsessively track the developments of a million AI systems I've never heard of week on week on week. What do you see coming now? Like, what feels to you like it is clearly on the horizon, but we're not quite prepared for it or won't feel until it's arrived? Maybe the way I'd put it is sometimes I've, and you've likely have the same, had the ability to have certain insights that have come through kind of reading a vast, vast amount of stuff for many different subjects and piecing it together in my head and having that experience of having a new idea and being creative. I think we underestimate just how quickly AI is going to be able to start doing that on an almost daily basis for us, going and reading vast tracks of human knowledge, synthesizing things.

coming up with ideas, telling us things about the world in real time that are basically unknowable today. The amazing part is people are going to have the ability to know things that are just wildly expensive or difficult to know today, or would take you a team of people to do. The sort of frightening part is, I think that knowledge is the most raw form of power. It's intensely like destabilizing to be an environment where suddenly everyone is like a mini-CIA in terms of their ability to gather information about the world. They'll do huge, amazing things of it, but surely there are going to be like crises that come about from this.

And I think the actual mental load of being a person interacting with these systems is going to be quite strange. I already find this where I'm like, am I keeping up with the ability of these systems to produce insights for me? Like how do I structure my life so I can take advantage of it? I'm very curious about how you think even having that ongoing conversation with the systems changes. you. So let me, I'll say it from my perspective. One thing I have noticed is that Claude is very, very, very smart. It is smarter than most people who know about a thing in any given thing. That is my experience of it. But it is not in the way that other people are, an independent entity that is rooted in its own concerns and intuitions and differences. What it is instead is a computer system trying to

adapt itself to what it thinks I want. So as I've talked to it much more about issues in my life, about issues in my work, various kind of intellectual inquiries or reporting inquiries where I'm trying to figure out questions that as of yet I'm at a sort of early stage of exploration, And what I've noticed over time is that one difference about in talking to it is it is always a yes and. Yep. It is never a no-but. It's never a honestly, are we still talking about this? It doesn't create, in the way that talking to my editor does, or talking to a friend does or my partner or anything.

It doesn't create the possibilities that another human does for kind of checking yourself. Yep. It's always pushing you further. And it's not necessarily bad. It doesn't always lead to psychosis or sycophancy or anything else. But it is very reinforcing of the eye. Yes. And I don't wonder about it so much for me,

although I actually even already feel the pressure of it on me. It's like, oh, like more good ideas coming for me, more interesting things I've come up with. But I do wonder about kids growing up on World War II, They always have systems like this around them. And the degree to which, you know, there is some amount of my communication with other human beings is now offloaded into communication with AI systems. I noticed that already being a kind of cage of my own intuitions, even as it allows me to run further with them than I maybe could otherwise. But I'm pretty well formed.

And you've got young kids, as I do. I'm curious how you think about what it means. how it will shape our personalities to be in these constant conversations. This is maybe my number one worry about all of this, is if you discover yourself in partnership with the AI system, you are uniquely vulnerable to all of the failures of that AI system. And not just failures, but the personality of the AI system will shape you. If you haven't, you know, I'm going to sound very Californian here,

even by I'm from England, it's soaked its way into my brain, brain, you have to know yourself and have done some work on yourself, I think, to be effective in being able to critique how this AI system gives you advice. And so for my kids, I'm going to encourage them to just have like a daily journaling practice from an extremely young age. Because my bed is that in the future, there will be kind of two types of people. There will be people who have co-created their personality through a back and forth with an AI. And some of that will just be weird. They will seem a little different to like regular people and there will maybe be problems that creep in because of that. And there will be people who have worked on understanding themselves

outside the bubble of technology and then bring that as context in with their interactions. And I think that latter type of person will do better. But ensuring that people do that is actually going to be hard. But don't you think the way people are going to discover themselves is with the technology? I think you were one of the first people who said to me, I should try keeping a journal in the systems. And I've done that on and off. And one thing it does is it makes it more interesting

to keep a journal because you have something reflecting back at you and picking out themes and so on. But the other thing it does is I feel it as a pull towards self-obsession. Because, you know, I drop in, you know, audio

record a journal entry and I drop it in. And all of a sudden I have this endlessly interested other system to tell me about me and it connects to something I said. Ezra, you're going through an amazing journey here. And I generally can't tell if it's a good thing or a bad thing. But, I mean, we already know from survey data that a lot of what people are doing on these systems is adjacent to therapy. Yes, but this to me is I think it will change how these systems get built.

It will change, I think, best practices that people have with these systems. And I think that we actually don't quite understand what this. interaction looks like, but it's extremely important to understand it. I mean, just to go back how, in the same way that you can get Claude to ask you questions to more clearly specify what you're trying to do, and that leads to a better outcome, I think we're going to need to build ways that these systems can try and elicit from the person, the actual problem they're trying to solve, rather than kind of go down a freewheeling path together, because in some cases, especially people that are kind of going through some kind of mental crisis. That is the exact moment when a friend would say,

this is nonsense. Like, you would not make any sense. Take a walk and, like, call me tomorrow. Or let's talk about a different subject. I don't think you're reasoning correctly about this. But AI systems will happily go along with you until they've affirmed a belief that may be wrong. And I think this is just a design problem and also will be a social problem that we have to contend with. And I just wonder how much it'll be a social force. I think we've given a lot of attention, and correctly so, to the places where it moved into psychosis or sort of strange AI human relationships, we're seeing it through its most extreme manifestations,

and those will become more widespread. I'm not saying they are not worth the attention. But for most people, it is just going to be a kind of a pressure. In the same way that being on Instagram, I think, makes people more vain. In the same way that we have become more capable of seeing ourselves in the third person, the mirror is a technology.

I mean, I think it's funny that the myth of narcissists, he's got to look in a pond. Yeah. Right? It was actually quite unusual to see yourself for much of human history. So when the mirrors came out, they were like, oh, this is going to lead to some issues. There's a lot of interesting research on how mirrors have changed us. And as somebody who believes in the sort of medium is a message thing, AI is a medium. And it will change us as we are in relationship to it, probably more so than other things, because it is this kind of relationship that has a,

a kind of mimicry of an actual relationship. Yes. I've used these AI systems to basically say, hey, I'm in conflict with someone at Anthropic. I'm really annoyed. Could you just ask me some questions about that person and how they're feeling to try and help me, I guess, like better think about the world from their perspective? And that's a case where I'm not using the technology

to kind of affirm my beliefs or show I'm in the right, but actually to help me just try and sit with, how is this other person experiencing this situation? And it's been profoundly helpful for then going and having the hard conflict conversation, sometimes even saying, well, I talked to Claude. You know, me and God came to the understanding you might be feeling this way. Do I have that right? And sometimes it's right, but sometimes when it's wrong,

it's really helpful for that other person to have seen me go through that exercise and empathy and spending time to try and understand them before coming into the conflict. Do you have strong views on how you want to parent in a world where AI is becoming more ubiquitous? Yes, I have the classic Californian technology executive view of not having that much technology around for children, but I was raised in that format as well. Like we had a computer in my dad's office. My dad would let me play on the computer and at some point he'd like say, Jack, you've had enough computers today. You're getting weird. And I'm not getting weird. No, no, you've got to let me. And he was like, see, being weird, get out. I think finding a way to like budget

it your child's time with technology has always been the work of parents and will continue to be. I recognize though that it's getting more ubiquitous and hard to escape. We have a smart TV. My toddler, she can watch Bluey and a couple of other shows,

but we haven't let her have unfettered access to the YouTube algorithm. It freaks me out. But I see her seeing the YouTube pain on the TV, and I know at some point we're going to have to have that conversation. So we're going to need to build, pretty heavy parental controls into this system.

We serve 18s and up today, but obviously kids are smart and are going to try and get onto this stuff. You're going to need to build a whole bunch of systems to kind of prevent children spending so much time with this. I think that's a good place to end. Always our final question. What are three books you'd recommend to the audience?

Ursula Legrin, The Wizard of Versi, was the first book I read. It's a book where magic comes from knowing the true name of things, and it's also a meditation on hubris, in this case of a person with thinking they can push magic very far. I read it now as a technologist. Uh-oh. Eric Hoffer, the true believer, which is a book on the nature of mass movements

and the psychology of what causes people to have strong beliefs, which I read because I think that we're AI technologists have strong beliefs and maybe part of a strong culture that includes the word cult, and so you need to understand the science and psychology behind that. And finally, a book called There Is No Antimometics Division by a writer with the name QNTM, which is about concepts that are in themselves information hazards where even thinking about them can be dangerous. And I always recommend it to people working on AI risk as a book adjacent to the things they worry about. Jack Clark, thank you very much. Thanks very much, Ezra.

This episode of Esauclancho is produced by Roland Hoop. Fact-checking by Michelle Harris with Kate Sinclair and Mary Marge Locker. Our senior audio engineer is Jeff Gelb, with additional mixing by Isaac Jones and Amund Sahota. Our executive producer is Claire Gordon. The show's production team also includes Annie Galvin, Marie Cassione, Marina King, Jack McCordock, Kristen Lynn, Emmett Kellecke, and Jan Cobol. Original music by Dan Powell and Pat McCusker. Audience Strategy by Christina Simulowski and Shannon Busta. The director of New York Times opinion audio is Annie Rose Strasser.