I'm Evan Goldstein. And I forgot the name of our podcast.
This is Open Questions.
Earlier this year, Jill Lepore, the Harvard historian and New Yorker writer, won the Pulitzer Prize for her 2025 book, We the People, a History of the US Constitution. She celebrated in what might seem like an unusual way, by stacking wood. In fact, this is entirely fitting given the polemical charge of her new book, The Rise and Fall of the Artificial State. Among other things, the book is a spirited defense of the natural world against the incursions of technological progress.
Evan, I know you're a Lepore fan. You've read a lot of Lepore, and you've interviewed her several times before in our pages. What intrigued you about this book?
In a sense, the topic is inescapable, right? Like, even if you wanted to think less about AI, AI imposes itself upon you. And so, someone of Lepore's intellect and stature taking up that topic is inherently intriguing to me, and there is no doubt that the intersection of AI, the meaning of college, how scholars do academic work, all are top of mind. And so, the opportunity to talk to her about some of these questions and get her responses was immediately appealing.
What really fascinated me about this book is the sort of history of science strand in it. There's a lot about how philosophers and inventors from Descartes all the way up to Sam Altman have thought about things like automation, consciousness, intelligence, and even like the soul. And so, the history of science stuff kind of immediately blends into a history of religion. One of the things that Lepore is really interesting on in this book is about how AI is a kind of magic. It has a kind of salvific religious force for a lot of its biggest proponents and maybe an apocalyptic and hellish force for some of its biggest opponents.
We of course also wanted to know whether she thinks it's going to destroy the academy. I was struck by a recent op-ed she wrote in the Harvard Crimson. And among other things, Lepore riffs actually on a title I think you came up with, Len, for a recent essay by the philosopher Anastasia Berg in our pages, and the essay was titled 'The Ivy League Surrender to AI Is Insane.' So Lepore takes that headline to basically pose a, I would say, equally provocative question, which is, has Harvard lost its mind? And for Lepore, the most recent cause for alarm was some remarks that really got a lot of attention across higher ed from Harvard College Dean David Deming, who suggested in a couple different venues that acceptance and even encouragement of AI in student writing was a logical policy position for the university.
So we didn't just talk about AI. We also talked about the prescience of Hannah Arendt, why the craziest stuff tends to happen on the most elite college campuses, and the anger of students that Lepore talked to across the country who feel not just that social media ruined their childhood but that now AI is ruining college. One little audio note: if you hear periodic rustling of animals and barking of dogs, it's because Lepore spoke to us from a farm, which, given her passionate defense of the natural world throughout this book, seems pretty appropriate. I'm Len Gutkin.
And this is Open Questions.
Jill Lepore, welcome to Open Questions.
Hey, thanks for having me, Evan.
Let's start big and broad here. What is the artificial state and what is it doing to us?
Well, we are building it, so we are doing it to ourselves. I wanted to give a name to what I think is unfolding, and I was kind of unhappy with the names that you come across in academic scholarship. My last book, We the People, which was a history of the US Constitution, was really a history of liberal constitutional democracies, of which the United States has been the most influential example. The epilogue of the book is called The Future of Constitutionalism, and I was really trying to figure out what kind of batons to hand to my reader to say, you know, okay, this is the past, what might you expect moving forward? And I have a great appreciation for the genius of the US Constitution and also a humble realization of its many deficiencies and inadequacies, especially in the present moment. And so the end of the book was really about my claim that an 18th-century constitutionalism that dates to the age of the printing press and the very beginnings of the Enlightenment idea of progress, which was tied less to technological progress than to moral progress, is insufficient for addressing the challenges to democratic societies that are presented by the destruction of the natural world and the rise of artificial beings. I there talked a little bit about what might replace the liberal democratic constitutional nation-state and began to describe the artificial state. So, if the liberal democratic nation-state is government by consent, this historically quite new way of people to live together, the artificial state is government by machine.
Are we living there yet, or is this a kind of unrealized or maybe only partially realized aspiration on the part of people like Elon Musk or Sam Altman and maybe Donald Trump?
Yeah, I mean, I describe it as something that's half-built. I also describe it as both a construct and a fantasy. I think it really is a fantasy. It's not aspirational for most people living on the planet right now. I do think it exists as a fantasy in the imagination of someone like Musk or someone like Altman. I mean, like Altman for instance, a couple of years ago said he thought an AI president would be better than a human president. It would be able to instantly and constantly determine the preferences of all of the population, of all of the electorate, and act on it. Which really is a system of rule that has nothing to do with our system of representative democracy and is just a terrifying notion, and is actually also a notion that has long appeared in science fiction stories that are dystopian. I'm not exaggerating to say it's a fantasy to these people. Like, they think that sounds great. That it has nothing to do with how we have chosen to live together, those of us who live in democratic societies.
So it's a fantasy of technologically enabled super efficiency that bypasses representative democratic channels. What are its tools? What are the actual mechanisms that somebody like Altman or Musk would hope to use to build and operate the artificial state?
First of all, I think it's important to kind of walk back imagining the listener who's like, 'Does she mean like the deep state? Like is it conspiracy?' And I don't mean that at all. I don't mean a conspiracy. I don't think this is secret. I'm trying to just describe what it is that I think is slowly being built and that has largely been built with good intentions. A lot of the ways in which our government bureaucracies operate using machinery is so that social welfare benefits can be more expeditiously delivered, so that we can keep track of the population for purpose of accurate proportionate representation in legislative bodies. We count people and turn people into data for all kinds of reasons. So veterans can get their benefits on time. So I don't mean to denounce the machinery itself, the infrastructure by which in the modern administrative state so much takes place. But a counter example of the kind of thing that I talk about as worrying and as something that is part of a future that I don't think anyone really asked for except for the people who run and own social media companies would be the public sphere. So since the rise of social media, which as you know has been entirely unregulated in the United States, most people's perception of what's going on in the world is mediated by social media platforms. I mean, many of us get our news from reputable news organizations that are doing tremendous and invaluable reporting or from public media, but the majority of people, let's sort of just confining ourselves to the United States, are getting their information from TikTok or Instagram in these kind of snippety ways and becoming what political scientists call low-information voters. And you might say, 'Well, okay, well, maybe, you know, they're just getting more information than they did previously when the news was on at 6:00.' There's plenty of evidence to counter that claim, but even if you were to stipulate that, at this point there's more bot traffic online and on every social media platform than human traffic. And those bots are owned by private corporations that have their own agendas, are often political agendas, are, you know, certainly financial economic agendas. They can be owned by foreign actors, they can be owned by terrorists, they can be owned by criminals. I mean, some of them can be benign, but it's clear that our public sphere is not actually a public sphere anymore.
So you have this wonderful quote where you say, 'Liberal democratic states make citizens, the artificial state makes trolls.' Trolls are associated with internet culture, they're kind of native to social media. So it seems like social media is one of the sort of primary specific mechanisms whereby the artificial state, in your view, supplants or seeks to supplant other kinds of state formation or state operation. And I'm wondering if you can just say a little more about how you see state actors using social media in that dangerous way.
When I say liberal democracies make citizens, like I'm referring to a body of political science scholarship on the ways in which our obligations as citizens are taught to us in the way that we live in civil societies. We agree to have our interests represented in proportion to our population. We agree to participate in forms of deliberation, everything from jury service to going to a, you know, town meeting or city council meeting. And we agree that it is an obligation of us to participate. So, participation, representation, and deliberation are those sort of three legs of citizenship in a democracy. And living in a democracy teaches you to do them, to, you know, to be aware of them. And the artificial state, as I say, just makes trolls. Like it kind of unmakes our capacity for citizenship by alienating us from one another, by removing from our observation our obligations as citizens. And even if you think about representation, I mean, the thing about social media is people are not actually agreeing to have their views represented in proportion to their population. People are agreeing to just like shout their own views, right? And then whoever is the loudest or the nastiest somehow is going to win that argument. Like it's sort of like a reversion to a kind of political barbarism. Like you may not be a lover of the 18th century Enlightenment and its ideals of the public sphere, but those were ideals and they involved an agreement of people, and in this case in the 18th century a small portion of the population, you know, to participate according to certain kind of rules of deliberation and engagement with an eye to the public interest. This is like a free-for-all out there at social media. And if that were just, you know, a form of entertainment or that was kind of what was also going on, that would be one thing. But when it supplants the actual public sphere, and maybe some of that, you know, was considerably worsened by the pandemic when people stopped going to city council meetings and would watch them on YouTube and then just [ __ ] about people in the comments. So, I don't really have a set of claims about the motives of state actors or non-state actors because they're not even really particularly subject to scrutiny. I don't have like a quantitative analysis of bots that are funded by people trying to sell lawn care products, what they're actually up to on social media, I don't know. But it's not a public sphere. And perplexingly to me, it came to be marketed as one. Think about Twitter, which was founded in 2006. Twitter was founded as a like a goofy fun thing, like just a lark. Post about do you like pickles in your tuna melts or not? Like just nothingness. But people who enjoy exercising influence over other people found Twitter to be a really interesting cudgel to use politically. And Twitter came to see that there was a lot of political content on Twitter. And by 2012, Twitter's publishing like the Twitter elections handbook and advertising itself as the new town hall meeting. You can get your pulse on the electorate. I think by that point like every member of Congress had a Twitter account. People are using Twitter to assess public opinion, to ask what their constituents might believe, to seek advice about how they might vote on a piece of legislation. And at that time, fewer than one in five Americans had a Twitter account. Most people who had a Twitter account just didn't even use it. The people who did use their Twitter accounts didn't really post about politics. They were still on and on about the tuna sandwiches and the pickles. And 90% of all political tweets were posted by fewer than 10% of users who were deeply politically engaged, highly partisan. And as the most deeply politically engaged people are, they tended also skewed to the left. They're highly educated, highly vocal, and politically extreme. And that minority of a minority of a minority, that's not even a representative minority. It's not like a representative selection that a like a scientific poll might produce. Twitter is advertising as a good way to sample the electorate. And then we're surprised at the exacerbation of polarization. I mean, to be clear, polarization in the United States started in the late 1960s, early 1970s, and we can't hold social media accountable for it, but from the vantage point of where we stand in 2026, 2012, 2014 of Twitter, that was the good old days, cuz at least it was humans. And now, you know, it's X. This is political violence has changed cuz Trump has kind of adjusted the dial to the right, but it's also mainly bots.
It seems like there's a certain ambiguity or maybe an ambivalence on your part in the book about, given that this is a kind of fully corrupted and corrupting sort of simulacrum of a public square that's doing measurable harm in some cases and doesn't seem to be doing much good, how heavily regulated should it be? You note that it's very unregulated here in the United States. On the other hand, the countries that do tend to try and regulate these platforms do not have governments that we admire or consider free. So, what's your take on regulation as a way of sort of curbing this aspect of the artificial state?
Here, I'm going to disappoint you. I am not a policy maker. I don't have a policy bone in my body. Like, what is the right thing to do here? Yeah, how much regulation is too much regulation? How little is too little? I don't have answers to that. I don't have experience legislating. I'm staying in my lane here. I don't think it's my obligation as a critic to present that. If I had what I thought was a great idea, I would have tested it on people who know a lot more than I do before trotting it out. I mean, I have a piece in the New Yorker that raises the question, is AI above the law? That traces out the history of why it is that AI is utterly lawless. And what I argue there and what I have argued elsewhere is that we have the internet we have, the social media that we have, and the AI that we have because Congress has completely abdicated its obligation to engage in any legislation whatsoever in any of these realms, and that's not an accident. Milton Friedman in the 1970s said, 'Hey, everything should be left to the free market. These environmental regulations, stuff like that, no. Regulation will stifle growth and stifle innovation.' And made that argument on television throughout the 1970s. And Reagan comes into office elevated by the conservative revolution in the United States and pursues a policy of deregulation that aligns very well with the emerging personal computer industry and the electronics industry in California. And the script is kind of already written by the time you get to 1984 and Apple rolls out the Macintosh computer that year with its incredibly great Super Bowl ad that year, 1984 will not be 1984. It's sort of Orwellian claim that the personal computer will inoculate us from totalitarianism. It'll be personally emancipatory that oh, we're devising these new tools and they're going to be great. And you know how they're going to be great is they're going to be good for democracy. I think it was kind of largely a joke. Like I don't really think people especially believed that about the personal computer. It was a Super Bowl ad, you know, it wasn't a treatise. But by the time you get to the 1990s and, you know, the Declaration of Independence of Cyberspace, that John Perry Barlow manifesto from 1996, libertarians have really won the day. The Telecommunications Act of 1996 is written by them and it says this will be a lawless space, which is exactly what the libertarians wanted. And that model is really well set up for social media. So when social media is rolled out, people are like this clearly is also going to be a lawless space. That it needs to be lawless in the interest of the First Amendment was really never any part of the claim. It's a kind of after the fact like oh, it's also going to be good for democracy claim, which is the same playbook that the AI companies have been using, which is perplexing because none of these really important and really exciting technological revolutions have actually been good for democracy. It's not that they couldn't be. They're not inherently bad for democracy, but the way they have been rolled out has not had the consequences that was promised.
But isn't it also the case, you talk about this in the book, that the tech companies, the social media companies, Facebook certainly would bend over backwards to regulate in compliance with, for instance, Chinese requests if they thought that that was what they had to do to open the market in China? I guess what I'm suggesting is there's nothing necessarily intrinsically sort of libertarian about these platforms or even about the ideologies of their founders. They're willing to compromise when they need to for market reasons.
I'm not sure I would agree with that altogether. I mean, first of all, they're different people, right? And Zuckerberg is a different person than Elon Musk, who's a different person than Sam Altman, who's a different person than Peter Thiel, who's a different person than Marc Andreessen. And it's understandable that we kind of glom them all together. There's so much that they have in common that they seem to not have in common with the rest of humanity. So much has been written and there have been so many whistleblowers on the Facebook stuff that I think we have a better view into the kind of lapses of ethics and the company over country motto that was Facebook's motto for a long time. And I don't think that we necessarily can see that as clearly with other people.
I want to get more deeply into your book by asking about Hannah Arendt, especially some of her writing at the dawn of the space age. She somewhat famously described the launch of Sputnik as second in importance to nothing, including the splitting of the atom. I wonder if you can speak to how Arendt has shaped your own thinking and writing at the dawn of the AI age.
Yeah, I read a lot of Arendt in graduate school and I hadn't really read that much since, but I had the great honor of being given the Hannah Arendt Prize for political thought. It's a German prize some years ago. So, I really, really read a lot a lot of Arendt. And it stuck with me. And so, last summer, I had to write the Tanner Lectures for Yale. I wanted to expand on this idea of the artificial state that I had earlier introduced. I went back and reread quite a bit of the Origins of Totalitarianism from 1951, which sort of ends with Arendt thinking about the relationship between totalitarianism and automation. And she makes the claim that they have something in common in that they both render humans superfluous, the first politically and the second technologically. And it seems that that insight contributed to her thinking and the thinking behind The Human Condition, which comes out in 1958 right after Sputnik. And the book really begins with Sputnik, you know, the first artificial satellite. Of course, artificial intelligence, the term is coined in 1955. Like, remember that the '50s, like, everything that was artificial was great. It was like the natural of the 1990s. Like, artificial sweetener, artificial nylons, artificial weather. Like, your air conditioning was artificial weather. You could sell something by declaring it to be artificial. And she was actually fascinated herself just by the preference for the artificial over the natural. But she begins The Human Condition reflecting on a reporter who, she says, described Sputnik as the first step in man's sort of long-awaited escape from the prison of the Earth. And her response to that was one of just really like, uh, astonishment. Whoever said it was a prison? Who's been waiting to escape all this time? Like, it's our home. It's a home to be tended. I wanted to be an astronaut well into when I was in college. I really, really am interested in space exploration. And I don't think Arendt had any problem with that as an idea, as the exploration of a new realm of knowledge about the universe. But what came with it, this notion of escape, this sort of reverence for the artificial and the denial of our animality, just I think kind of like creeped her the [ __ ] out. I mean, honestly, like, I think she was really creeped out by it. I mean, certainly I am. And it was kind of mystifying. I think a lot of people sort of feel mystified about a lot of stuff that people in Silicon Valley are like, 'Really? Like, you think that? Like, that seems like a good way to live to you?' Yeah, I had that in mind when I was thinking about the parallel between the Sputnik moment, it just kind of changes everything, and the ChatGPT moment in 2022, which was immensely exciting. Like, I am hugely excited about artificial intelligence and its many possibilities for research. But then its sort of reception relieves us of the obligation to talk to one another, to thank, to write, to code, to do math. I don't know, like what? That's my like creeped out response. So, it seemed like a good place to begin the book.