Dr. Jill Lepore on why AI backlash is vital for the future

The Artificial State author explains how counting everything breaks society — and how college grads booing AI might save us all.

Today, I’m talking with Harvard professor and New Yorker staff writer Dr. Jill Lepore about her new book, The Rise and Fall of the Artificial State , which comes out on August 25.

Jill is one of the best writers there is at identifying institutional patterns in history, and Decoder is a show about systems, so this book could not possibly be more in our zone.

You build systems by measuring things — gathering information, quantifying it as data, and taking action on it. And Jill’s big thesis in the book is that this kind of systemic measurement is really a dehumanizing act, one that flattens human concerns into something, well, artificial.

These artificial structures touch everything — not just in the more obvious places like in apps or AI, but all the way up to politics and governance itself, and Jill’s insights here are a telling portrait of the way Big Tech and its billionaire figureheads have managed to break the very idea of the state itself.

You’ll hear Jill tell me she wanted to write a “kind of grumpy book,” but in my reading, it’s a pretty dead-ahead look at the last century and a half of technological and social developments and how we got to a place where bots are maybe running everything.

Jill’s idea of the artificial state starts with the very idea of quantification, from the first stabs at taking a census thousands of years ago to the invention of modern political polling in the 1930s. But where she really sees a big pivot — and the era she and I talked about most — is in the 1980s, when we saw the emergence of the Silicon Valley culture we have today, when computers got personal, and when computer-aided targeting and microtargeting really truly became both possible and commonplace, both for ads and for political campaigns.

What Jill and I did disagree about is some of her pessimism about the internet, because the internet is what made jobs and careers like mine — and businesses like The Verge , and shows like Decoder — even possible at all. There really was a great democratizing of access that happened in the late 20th century and into the 21st, as people got online, and not all of it was as illusory as maybe the great hopes for Twitter after the Arab Spring turned out to be.

But as I said, Jill thinks in systems, and so her response, about the way systems of power keep adapting to how people use new technology, was as illuminating as ever.

Of course, we also talked about the future — including where she thinks she still sees hope for the future. And yes, that does include college kids booing the AI CEOs at graduation ceremonies.

Okay: Dr Jill Lepore, on the Artificial State. Here we go.

This interview has been lightly edited for length and clarity.

Dr. Jill Lepore, you’re a staff writer at The New Yorker , you’re a professor of history and law at Harvard University, and most importantly, you’re the author of a new book, The Rise and Fall of the Artificial State . Welcome to Decoder .

I am really excited to talk to you. I’m a big fan of your work over the years. And I have to say, this book is one of the most Decoder -y, Verge -y, Nilay-Patel-strike-zone books of all time.

[Laughs] Wow. I’m not sure how to take that, but I guess that’s good.

The producers and I just had a thrill ride reading this book. It is essentially about how computers and measurement flatten us and then systems can take advantage of that flattening. Like I said, it’s in the strike zone.

So let’s just start at the very beginning for folks. The book is called The Rise and Fall of the Artificial State . Tell us what the “artificial state” is. Define it so we can start from that foundation.

What I mean by the artificial state is a successor to the liberal nation-state, the constitutional democracy under which our freedom of speech, free discourse, and constitutional rights are central to our experience as citizens within a nation-state.

We talk a lot about democratic decay and democratic backsliding, and there’s a whole public discussion of what’s happening with democracy, but I don’t think we talk about the nation-state as much as we might. And that is to say the state itself.

And so what I mean by the ‘artificial state’ is a state in which more and more of the functions of the nation-state have been taken over by private multinational corporations. The easiest example of that to think about is what is euphemistically called the public square, which is where people get information about how to vote, what’s going on in the world, how to make decisions, and how to exercise their rights and duties as citizens.

Our public discourse is almost entirely privatized and owned by corporations. As many, many people have demonstrated — it’s not news to your listeners or to anyone — the priority of public discourse is not the free expression of ideas and democratic deliberation, it is the monetization of attention and of outrage.

More than that, a fairly recent vintage is the fact that public platforms in which we exchange ideas as citizens and democracies are dominated by bots. All of them have been inverted for several years now, or different numbers of years, from X to Facebook to TikTok. There are far more bots than humans on those platforms.

Many of those bots are owned by corporations and have particular acts, their own kind of financial interests, but they’re owned by state actors, non-state actors, criminals, and terrorists. What we’re at risk of is losing the liberal constitutional democracy and having it become supplanted by what I describe as the artificial state.

One of the things that really struck me about this is at least America’s conception of itself, the United States’ conception of itself is this messy, multi-part democracy. The Senate loves to call itself the most deliberative body in the history of the world, and it’s supposed to be a whole bunch of people that are ruled by the consent of the governed and we’re all arguing.

Your thesis is that that’s what we say about ourselves, but actually we’ve automated away and numerically quantified away a bunch of that, and we are now run maybe by bots on Twitter, or at least politicians are overwhelmed by bots on Twitter and they’ve outsourced their thinking, or there are consultants that are turning us all into polling data. Is that the trajectory we’re on that ends the nation-state? Is that the core of the idea?

Yeah. It has a long history, but the acceleration of these developments in the age of artificial intelligence is really considerable. I’ll give you two examples that stretch out across a century. In the 1930s, we see the invention of so-called scientific polling. George Gallup starts the American Institute of Public Opinion research in 1935.

What polling does gradually over the middle decades of the 20th century is replace the kind of door-to-door precinct walking that campaigns would need to do, right? If I’m running for election in some neighborhood, I have to go talk to people. I’m going to go to baseball games, basketball games, and the public library. But increasingly, it became less necessary for political campaigns to do that because they could subcontract out political polling.

You can find out what political opinion is in a neighborhood if you want to run a campaign just by doing data mining and data analysis.

We might say, “That makes politics so much more efficient. Those campaigns can more efficiently use their money and they can get their message out to the right people.” But if you think about each one of those conversations that people used to have — “I’m a campaign volunteer and I’m going to go and I’m going to sit in on some PTA meeting,” or, “I’m going to go to a basketball game in some neighborhood and talk to the parents who are there” — all those conversations are the civic glue that holds a civil society together.

Replacing that chatting under a street lamp about, “What do you think about this guy?” with just extracting data, which is where we are now, is a much more artificial politics than we have had before.

There’s been great coverage lately of emerging trends for the 2026 election in the US,and around the world, where a lot of people report that even in these primaries that have been going on over the spring and summer, they’ll ask ChatGPT, “How should I vote in the primary?” Meanwhile, many of the campaign messages that you’re getting have been crafted by artificial intelligence. The fact that you are getting them is a decision made by artificial intelligence. In a way, it’s AI versus AI in these upcoming elections, or even in elections that are already transpiring.

You can see how you could fall into that step, just going downhill on skis. But I wanted to write a grumpy book to say, “Let’s pause for a minute. What’s at the bottom of that hill? Is that an icy pond and we’re going to just go through the ice and we’re all going to drown in there? Or is that a beautiful new political world?” And I think it looks pretty perilous.

The book traces this evolution step by step as you’re describing; this is a slow build to maybe where we’ve gotten today.

As you point out several times in the book, there are a bunch of science fiction authors in the ‘50s and ‘60s who were looking at computers and drawing a picture of the future. Some of them picked utopia, and there are entire movements about technocratic utopia that emerge, and a lot of them pick dystopia. They draw a picture of computer-aided dystopia. You point this out. Everyone has heard the Torment Nexus meme. It’s in the book. You use it in the book. Our modern industrialists have read dystopian science fiction and decided that is the model of the future that they’re attempting to build.

I’m curious why you think they’ve all gotten it wrong, from Elon Musk to Peter Thiel on down, and if that is an inflection point for us to actually pick a different future.

It’s a big thing, starting in the ‘30s, to imagine a future in which the president of the United States is chosen by a machine.

There are a bunch of examples of this from the ‘30s, where people are really worried about totalitarianism and the dehumanization of people as a consequence of the machine age. But there’s also a lot of this in the 1950s. For instance, a classic story by Isaac Asimov that I think is from 1955, called “Franchise,” and it’s set on the days before November 4th, 2008, the day Obama won the presidency.

In that imagined future, there’s this machine called the Multivac that determines, using polling data, who in the United States is the average man, the most representative man. The sampling method that pollsters use has been perfected mathematically by this computer to pick the average person. And that’s the only person who votes. And then that guy has to go into hiding because if people don’t like who was elected or the presidency goes badly, his life is destroyed.

There’s a lot of this stuff in science fiction, but Sam Altman was on Joe Rogan and he started talking about how great it would be if AI chose the president of the United States. Yes, he was fucking around, but he was also serious. “Wouldn’t that be great? Because it’s so efficient and we don’t have to bother and it spares us the labor.”

And that general premise, that “Why would we bother to do something that a machine can do for us?” without the realization that the rest of us would have. There are some things we want to do, including meaningful work: writing, making art, and doing things for ourselves.

I know these are very smart people, but they’re not humanists. Thiel would say that he’s a political philosopher. I’m not so convinced of that. The capacity for humanistic inquiry, reading, and reasoning is not something you see in great evidence in that culture.

It’s been bred out, in a way, of the culture of Silicon Valley. Even if you look back to the ‘90s and the culture of the celebration of the serial entrepreneur as the new man of progress, that particular kind of cultural formulation was all about the disavowal of the past.

It doesn’t matter what came before. You want to start a ride service, you don’t study taxis or figure out where delivery service came from. You don’t look at traffic patterns. You don’t study anything in the past, you just start new. You just imagine it all new and you destroy what came before. There’s such a rejection of the knowledge and wisdom of earlier human beings.

The posture of the humanist is one of humility. Isn’t it magisterial and majestic how Homer wrote The Odyssey so long ago and it’s still so powerful? The timelessness of literature and the arts, the infinite beauty of humanity’s capacity for making things, it’s just not the ethos. It’s, “No, we’ll do it better. It doesn’t matter what anybody did before.” It’s, “We’re innovating, we’re disrupting.”

The culture so fetishizes the future that it doesn’t have that, “Let’s listen to what other people who came before might have known, that’s still wisdom.” That makes for a very poor reading of literature. We could have a great argument about what Isaac Asimov meant by that story. It’s not a determinant, I’m not the authority on Isaac Asimov, but I’m not going to ask Sam Altman what a short story means.

There’s something very interesting about this cast of characters, particularly Peter Thiel who comes up in your book, Elon Musk who comes up in your book somewhere over 100 times I think. They are falling back to literally Catholicism as some kind of moral foundation for whatever decisions they’re making. And whether or not they’re misreading Catholicism is irrelevant to them. They’re often in fights with the Pope , which is just not something a straight reading of Catholicism would allow you to do.

There’s an entire culture of thinking — I won’t call it a philosophy, because I don’t think it’s quite that developed — around first principles. “We’re going to do first principles thinking and that will lead us to some disruptive outcome.” It often, to me at least, feels like where that first principle begins is a choice about what to ignore from the past.

I see that over again in your book, in the culture we live in at large, that deciding what to set aside is the big decision. That’s an assumption that is often not questioned as we go step by step towards a more quantified society, a more automated society.

But there’s something messy about democracy that is very, very important. The decision we’re making now is to actually set that mess aside in the name of, I don’t know, efficiency, potentially happiness. I’m curious if you see that setting-aside is happening by accident or on purpose?

I think of these guys as really instrumental. They’re not pursuing knowledge for its own sake. They are deeply interested in exploration and wonder, and one of the things I really love about this moment is the fascination with space and space exploration, which I think is tremendously exciting. Harnessed to militarization in certain commercial projects is like, “Wow, we went from Star Trek to Star Wars really fast.”

But the sense that what you might want to know about the world can be just because it’s worth knowing rather than that you can use it to extract money from other people, I just don’t think that’s part of that tech culture. It’s just not. A humanist commitment to knowledge for the sake of the accumulation of wisdom, leading a better life, and understanding the world with greater clarity and greater compassion is largely mocked in that world as “woke” or “feminine” or “libtard,” whatever. Especially with the conservative turn of the tech world, it takes on weird valances.

It is really purposeful. “This kind of knowledge is not worth having [because] we can’t make money off of it. If we had that kind of knowledge, it might slow down what we’re trying to do. And what we’re trying to do is win the AI race and determine the future of humanity.” The lack of curiosity about the bigger world and the place of these really thrilling inventions within it, I don’t know, I find it bizarre. This causes a lot of self-reflection on my part. A lot of these people are my former students. I teach at Harvard. We send so many students off to work in Silicon Valley, and then they talk this way and what happened? What did we do wrong? Do you not have a sense of obligation to the human condition?

You have a number of these markers. Walk us through some of them, from “the internet and social platforms will cause an outbreak of democracy” to where we are now in 2026.

That first began to emerge in the 1980s. If you go back to 1984 and the launch of the Apple Macintosh and that famous, iconic Apple TV ad from the Super Bowl in January of 1984, the promise was that the personal computer would be revolutionary. As I’m sure you’ve talked about in your show and your audience really understands, the celebration of the personal computer as emancipation came out of the counterculture. There’s this great book, From Counterculture to Cyberculture , that’s about the so-called “California ideology” that marries counterculture libertarianism to computer utopianism.

And you see that realized in 1984 with the launch of the Macintosh: “We’re going to defeat totalitarianism in the form of the giant mainframe IBM” that’s associated with two things at the time — big businesses, which are the only places that can afford these giant IBM mainframes that are running payroll systems and stuff for big corporations, and the Republican Party, which actually has a ton of IBM mainframes. The Democrats have basically no computers. It’s a huge problem for the Democrats in the 1970s and 1980s. So there’s this whole feeling that the personal computer is going to advance liberty, personal freedom, and therefore democracy. It’s a libertarian vision, but it’s about emancipation. And it’s an ad. At the end of the day, it’s just an ad campaign.

There were people who said, “Well, exactly how does the personal computer emancipate? It’s great. I would like one. It’s fun to have. I also like a nice toaster, but my toaster isn’t going to emancipate me. What is it about?” You begin to see new tech computing products as very carefully packaged with an ideal about personal liberation. By the time you get to the ‘90s, and the opening of the internet to commercial traffic, you see the scaling of that promise. “Now we all are going to have personal communities that are going to be linked to one another, and so society is going to become emancipated and democratized. We’re going to have a digital democracy.” Even though it’s a little bit like when political pundits make election forecasts and they’re wrong half the time, but no one ever remembers even what they forecasted. How do you even have a job?

The people in the ‘80s said the personal computer was going to liberate people. Well, it didn’t. But then in the ‘90s, the internet was going to democratize societies. Well, it didn’t. That was just an ad campaign. And not that there aren’t plenty of people that sincerely believe that; there certainly is tremendous democratizing potential for the internet. There was a lot of promise there, but it’s not realized for all kinds of reasons.

That’s already two strikes, but then you get to 2004 and all the emerging social media says, “Oh, Facebook, we’re going to democratize this. We’re going to connect everybody. Everyone will be able to be close to one another.” That also doesn’t happen.

Then the ad campaign now, the packaging of somehow AI is going to do this. “What we need is democratic AI.” And that doesn’t mean we as a polity have a say, it just means that democratic countries should lift all restrictions and regulations so that democracies can have AI before authoritarians. It’s an old story.

Can I push back on this a little bit? Because I personally struggle with that ad campaign, but the reality of it. There’s not a world previous to the internet where a guy like me who grew up in Wisconsin with no connections to the media ends up interviewing you, a Harvard professor. The internet needed to arrive. Publishing needed to be democratized. For whatever reason, AOL needed to pay a bunch of bloggers $12 a blog post, and then somehow decide that I should be one of those people. A lot of things had to happen where gatekeepers did fall away, and people like me were allowed to go find their audience. And I see it every single day now with young people who style themselves as creators or YouTubers, that they know that they can start a business and reach people on these platforms in some way.

The flip side of that is that I do think a bunch of gatekeepers, the men who own these platforms — it’s largely men — realized they could turn the knobs. Elon Musk could buy X and he could take X from a platform where, when it was known as Twitter, lots and lots of activists could find each other and shape public opinion. And now X is whatever X is, and mostly racists find me. That’s my experience of that platform whenever I participate.

Mark Zuckerberg started a Supreme Court for Facebook moderation because he wanted to recapitulate the machinations of democracy for people who had been banned for being crazy in their local Facebook groups. Maybe that worked and maybe it didn’t. My friend Casey Newton’s going to have something to write about in Platformer for the rest of his life as that experiment continues. But Zuckerberg said, “This is kind of dumb. I’m turning it off. We’re doing community notes here. No more moderation.”

There’s some kind of pendulum swing. I agree with you, this is largely an ad campaign, but a lot of people were able to find each other. A lot of new people were allowed to gain a much larger platform with which to affect the culture, if not society at large. How do you put that in the context of the artificial state? I’ll just preview it. My next question is going to be about targeted advertising, which I think is an enormously influential piece of the puzzle. But it’s that dynamic that I struggle with, it’s not purely an ad campaign.

I’ll come at that in a couple of different ways while granting that you absolutely have a point. I’ll figure out where we could maybe see a line. We tend to forget that, for instance, up until fairly recent history, maybe the 1980s, very few newspaper reporters in the United States had college degrees. They were people, mostly men, but there were a lot of women beat reporters and women editorial writers who started out delivering newspapers, working at the press, working the printer presses at newspapers, climbing their way up. I mean, there’s a reason Jimmy Olson is in the newsroom with Clark Kent in the 1930s comic books. They’re street kids and they know the streets and they start reporting. I don’t think they’re that different from coming up with a $12 blog post payment.

That’s what happens in every industry, unsurprisingly. A new medium emerges, the daily paper or the twice-daily paper. The interview is invented. Eyewitness reporting is invented, the byline is invented. The history of the newspaper is all about an explosion of new kinds of doors that different people can walk through to report on the world to other people, and then more people can learn about the world because these newspapers are different.

There’s an explosion in the democratization of any new technology of communication when it starts. By definition, if you can make a technology of communication cheaper and faster, you’re going to reach more people and more people are going to be involved in producing it. That happens with the first weekly newspapers, with the penny press in the 1830s, with the telegraph in the 1840s. It happens with how the telephone changes news reporting, the wire service in the Mexican-American War. It happens with the radio in the 1920s. It happens with TV in the 1950s.

In fact, what’s interesting about that is that each of those transformations reshapes the party system. There have been seven party systems in American history and they all turn on a new emerging communications technology, which unsettles the relationship between the elites and the people. The people get more power. The parties just have to get realigned.

But what always happens is the elites regain power. They don’t like it. They don’t like the news boys becoming beat reporters, becoming editors. They say, “Who are these nouveau people?” To use your metaphor, they turn the dial, they regain control, they consolidate their power over the medium, and then it has to wait for the next communications technology revolution.

The tricky thing about the moment that we’re in now is the emerging technologies are not restoring power to the people. They’re not turning the wheel this way, they just keep going that way. I teach undergraduates. I see them go out in the world. There are no fucking jobs for them! You want to be a writer, you want to be an artist, you want to be a movie maker, you want to be a screenwriter, where are you going to get a job? The whole culture industry is imploding. I don’t think this is a magnificent time to be out there being an influencer. You still don’t have health insurance. You still can’t buy a house. I just don’t think we’re in a moment of emancipation that’s responding to that consolidation.

Some people are like, “It’s great. I like that these billionaires have all this power over our culture.” I don’t, but we’re continuing to rotate in their direction. I don’t think it’s turning back. That’s how I would grant for you that there’s always something thrilling, emancipatory, illuminating, increasingly inclusive about a new communications technology. I wouldn’t deny that, and so you’re right to that degree, it’s not just an ad campaign, but I do think we’re in an exceptional moment right now.

I said I was going to ask about targeted advertising, and one of the reasons I think the pendulum has swung back so sharply is a reaction to what you might call identity politics. In the book, you say that identity politics as we know it is made possible by computer-run sorting and targeting. The algorithms on these platforms will put us all together, and then we will say, “Here’s an identity. We’re a bloc. We have some political power as a bloc. Deal with us.”

That seems true. I think we’ve all had that experience. We’ve seen that play out both in the culture, in our politics more broadly, but you can push back on that. You can say, “No, we don’t like that. Identity politics bad. That’s too woke,” and you can see the culture react to that. You can certainly see politicians and billionaires react to that.

But I don’t think identity politics is only a product of quantifiability, of computer-aided targeting and sorting. It certainly existed before. Do you think the identity politics of today is defined by the fact that we all exist on platforms, or do you think it’s just shaped by it? What’s the relationship there?

I remember being quite struck reading an article from 1968 in a magazine that I think was called Science and Technology , a popular magazine, like the Wired of the 1960s, by this guy, Ithiel de Sola Pool, who was an MIT political scientist..

In 1968, he was asked [alongside] JCR Licklider, the inventor of the internet, to speculate on what the computer revolution would mean by the year 2000. They were both really smart guys and they had really, really fascinating things to say. Ithiel de Sola Pool, this political scientist, said, “Certainly by then there will be personal newspapers. You will log onto your computer in the morning, and it will be networked,” because the computer was moving in 1968 from being a storage and data processing device to being a communications device, being both of those things. So he said, “You will be able to log onto your computer in the morning and be a personal computer.”

And Pool said, “There really won’t be able to be party politics if you have your own newspaper because a political party is the organization of political dissent, of shared political priorities. I don’t think we’ll even have interest group politics,” like the Black Congressional Caucus or the Corn Farmers Association. “I don’t think we’ll even have that. We’ll really only be able to have personal politics. I just think that our notion of a representative government where we agree to elect people to represent not my individual interests, but the interests of my community, we won’t be able to do that because no one will have a sense of what is in the interest of their community. You’ll only have the sense of your own interests, and that’s not how a representative democracy works. Our system of government will be incompatible with our technology of communication.”

Smart people saw this 50-60 years ago, and I think smart people see it now happening all the time. It’s something that is complicated for political campaigns to work with. You read interviews now with campaign consultants who say, “I really do not want to use these tools. I really do not want to do micro-targeted messaging, voter by voter, with deep-fake AI ads to convince people to vote for my candidate. I really believe in my candidate, but I don’t want to use these tools. But I have to use these tools because there’s no moratorium on using these tools, and therefore, I have to use the tools that I think my enemy will also use.” It’s a sad moment. I don’t think there’s anybody — well, I don’t know, there are probably some wretched human beings out there really excited to do deep-fakes and win elections that way, but most people in politics are not loving this moment.

I said I was going to ask about targeted advertising, and I started with identity politics for a reason. If I look at how money on the internet is made broadly today, it is targeted advertising. I’m just thinking about your line about identity politics and how it’s made possible by computer-run sorting and targeting. Computer-run sorting and targeting is advertising. That is the modern advertising industry at scale. “We will find you, the person who is most susceptible to whatever advertising message we have, we might know so much about you that we can target that message even more tightly to you in an automated way, and we will deliver it to you.”

I’m wondering how you see that connecting to the artificial state. We’re going to measure everybody, quantify everybody, we’ll do capitalism at scale, and now we’re going to subsume our politics into it. Because you’re saying that the politicians themselves understand that maybe they shouldn’t do that.

It’s helpful to remember that that’s not an invention of the internet or the personal computer era, or certainly not of the artificial intelligence era. There is a great book by my colleague, Liz Cohen, Consumers’ Republic , that argues that the consumer replaced the citizen as the individual entity of consequence in the United States in the 1950s, in the postwar era with the peacetime conversion of wartime factories. Consumption was what was going to save the country and advance the country’s interests, and American growth and American strength was associated with consumerism. In my lifetime, that’s been the way that we live. That’s not where the country was founded, and that’s also not inevitable.

But the system that you’re describing with targeted advertising, which is effectively what the internet currently is, I actually don’t think that’s going to last. We’re not acting online primarily as citizens. Certainly I don’t feel that I am. I’m chiefly acting online as a consumer. I remember the day that I noticed that. In like 2000 I had a baby, and Google started sending me ads for breast pumps. I said, “Fuck this.” Remember when you were outraged by that and you say, “I hate that they know that I had a baby. That makes me mad. I didn’t tell them that”? But now we just accept it..

Everyone’s phones are listening to them and it’s whatever.

Everyone’s phones are listening to them. You’re getting on a flight, you know that someone’s going to offer you Dramamine online. But that seemed, to a lot of people, inevitable and almost natural because the consumers’ republic is how we’ve lived for such a long time.

I’m not trying to offer up a Luddite or nostalgic reading here, but when you walk down a city street, there’s City Hall and there’s the post office, and — this is going to make me sound like I’m 112 years old — there’s the JCPenney and the Walgreens. There are different places. It’s now just the one place where the information that City Hall might have about you or the information you might wish to share at City Hall, your house has been reappraised and you want to question that, or you’ve got a parking ticket and you want to appeal it, those were in one place. And then if you’re going to go buy ibuprofen or new socks, that was a different place. It’s now all corporatized and privatized. But the privatization of the public square is a transformation that is a long time coming.

I’d be very curious to hear what you think about this, but the people that I read increasingly say the internet is just going to be for bots in fairly short order and the humans aren’t really going to be on it. It’s just going to be agentic AI doing your stuff. It’s really just going to be machines talking to machines, which is in a way what people thought it was going to be to begin with. I think that actually offers some prospect for restoring the real world as a distinctive place for how we engage with one another as families, communities, and governments. But where we are now, if this continues, I don’t see how democracy endures the corporatization of all public discourse.

We have had many, many, many tech CEOs on the show and we have asked them very directly , “What happens when no one uses your app anymore, when it’s just the bots and the agents doing my bidding, and all of your business model has to fall away, and you are just a provider of Toyota Camrys?”

Yeah, yeah! What do they say? What do they say?!

It breaks down in two ways. Anybody who offers a physical service says, “Yeah, it was really hard to build this business. Go ahead and try to disrupt me with your bots. You won’t.“The CEO of Zocdoc , the CEO of Uber , and the CEO of TaskRabbit all say, “Yeah, go ahead and try. Try to find an army of people to come hang up your TV on demand. That’s our database, that’s our labor supply, and we will not be disrupted, because you can’t replace us.”

Anybody who runs a virtual business or something that’s more software is freaking out. Amazon is basically freaking out. They’re a front end to a bunch of drop-shippers. They can be replaced. They’re the ones suing Perplexity . And Perplexity is saying, “Our software is the new labor, and you’re discriminating against our labor,” which I’m still trying to puzzle out.

Do I believe that the agentic AI is equivalent to my labor? Should it have the same rights as me? That is an open question that I think the internet is not quite ready for, but it seems like the bleeding edge of what it actually means to even have an internet. If I can deploy bots out onto the internet on my behalf, should they have the same rights as I do, or do you actually need me and all the friction that I come with in order to build a business model, and do you have a right to that business model? I’m wondering if you know the answer, because that’d be great for me. I could just skip to the end.

Yeah, I totally don’t know the answer. I’m fascinated by every time I look at the height of the tidal wave of the coming techlash, it’s higher and higher. I don’t see that manifesting itself in an effective way, but we’re in for really bumpy times.

The other piece of it that strikes me is all of “the internet will go away” is wishful thinking on a part of the companies. They want this internet to go away. The economics of this internet are really well-known to them. There are a bunch of gatekeepers that just extract fees at every turn of this internet.

Most famously there’s Apple. I think Elon Musk and Mark Zuckerberg and a whole bunch of other people would like to have business models that are not restricted by the iOS App Store and Apple’s 30 percent fees and however they feel about privacy. Saying, “We can build a new internet that’s all bots,” gets you away from the iPhone, gets you away from this device where Apple gets to decide what the internet is for in big and meaningful ways.

But I think people like Instagram, and I don’t think they’re going to give up their phones. That keeps you in a world where there’s at least something happening, where you’re looking for other people in some way. And that to me is maybe the heart of the solution.

I’ll connect it to the fact that you’re a college professor. You’re at the height of the techlash, as you just said. I’m watching college students boo AI at graduation ceremony after graduation ceremony and talk about it on the social platforms in big and meaningful ways. That seems like if there’s any kind of market pushback that leads to a political pushback that leads to actual change, that’s where it comes from, and that still came from the internet itself.

No, it came from those people in a quad sweating in their stupid caps and gowns making a lot of noise together. It came from an actual group of humans in physical space together. I wouldn’t look to college or public school administrations to be of any help because they’re all just saying, “Oh, we made a deal with OpenAI. Here’s your not-so-free tool you have to use now.” Institutions, not just individuals, have to have positions of moral integrity in this moment, and I don’t see a ton of evidence of that. You see it maybe in the publishing industry, it’s very self-interested, and I think you see it in some state governments and some local governments, but I don’t think you see it in big law firms, big universities.

I gave these lectures at Yale last year that are the basis for this book on the artificial state and had a lot of group meetings with Yale students. I had been told by a colleague that at the very first meeting of the AI task force, which the faculty thought was going to be about what our pedagogical position should be about — [was instead about] “We’re going to use this task force to deliberate on what’s the proper time and place for using these incredibly powerful tools, and then what places are maybe not good places.”

He said, “The first meeting of the task force was about how we can brand AI at Yale so that it’s really sexy and people are excited about it. Are we going to call it YaAI? What’s the branding strategy here?” The faculty said, “What are you actually talking about? You brought us here to have a branding conversation?”

I’m in a lecture hall with this wonderful group of students. I said, “Let’s just have a good discussion now. What about just having there be a two-week moratorium starting tomorrow? The president of the university could announce it tonight that, for two weeks, every class on campus is going to be a discussion section involving staff, faculty, and students on what our community policy should be about using this tool this year.”

The students said, “Let’s do that. Why can’t we do that? Why can’t we have a discussion about it?” The staff and the faculty said, “Why are we not doing that?” You say, “Why are we not doing that?” That’s why the students are mad. They’re paying all this money, taking on all this debt, working their asses off to be replaced by machines and have no capacity, be granted no fulfillment of the role of student, scholar, citizen, member of a community, and what that means and should mean.

It’s such a mismatch of the Silicon Valley rhetoric of the future of humanity. “This is bigger than the wheel or the invention of fire,” [against] like, “But we don’t give a shit what you think or want.” That’s what gives me hope, that restlessness around that. There was a resounding applause around this idea, and you will notice this is not what Yale did. Yale instead branded its relationship with OpenAI, and now it’s a corporate agreement.

Let me connect that sense, that lack of agency that you’re describing, to what we see in our politics writ large. I think a lot of people are going to their local town/community planning meetings where a data center is being proposed. They’re saying no, and then it’s happening anyway. We see this playing out. Little “D” democratic politics are failing people in obvious ways. “Here’s this data center that you don’t want, and you’ve literally come to the meeting to register your opposition. Well, we’re doing it anyway because the town needs the jobs,” or whatever excuse is being put forward.

We see this maybe most notably in Utah. Kevin O’Leary is going to build a data center no matter how much those people in Utah want it or not. And maybe the elected officials get thrown out because they said yes, and that will be the mechanism of change there. But actually, the representatives are not doing what the people want. The people feel this lack of agency.

Your entire book is about the decaying of the liberal nation state. I could point to that very tiny thing. People don’t feel agency about data centers, and say, “That’s the tip of the spear. That’s what’s going to break it wide open.” If you don’t believe in your local government and you haven’t believed in your national or state government for a long time, what’s left? Is that recoverable? The thing that I’m worried about the most is maybe that’s not recoverable. If you can’t even keep the data center from showing up in your backyard, what else is there to believe in? If you can’t even get Yale to take a beat and talk to you, what else is there?

Yeah. I don’t know, man. That is really grim.

No, no, you’re supposed to depress me . It’s the other way around. [Laughs]

I love that people are showing up and having those arguments. And if you were a smart community development board, you’d get ahead of that and say, “There are some good arguments for this data center that are honest arguments, and then there are some real arguments against it. Let’s have an open forum, and let’s see what the people think. And let’s also educate ourselves and the community,” because there’s a lot of dumbass stuff out there about data centers, I’m sure, certainly on both sides.

I don’t know. I think the fact that that’s happening and is likely to really influence these elections is meaningful. There are political consequences for just going along to get along. But it’s part of a much bigger problem with money and American politics and the relationship between corporations and government, so there’s no really easy fix for that.

To your question, structurally, is there any going back? My last book was about the history of the US Constitution. And one of the claims I make in this book is that it’s a problem for American politics and government when the Constitution becomes unamendable. The US Constitution hasn’t been meaningfully amended since 1971 . Instead, the Supreme Court amends the Constitution in how it acts. So the longer that goes on, the more resentful people are of the Supreme Court.

When people ask, “Would we ever get that back?” the example I often give is jury service. One thing that people really believe in is jury duty. Even if they don’t like to do it, they believe in it. They think it’s a good thing to have, that it involves fairness in our criminal justice system. Even if, objectively, juries often make the wrong decision, people believe it’s good that we are asked to go to a room with strangers and deliberate over what’s true and what’s not true.

But imagine that, for some reason, jury service was abolished in 1971. And now in 2026, someone said, “You know what would actually be good would be bringing back jury duty, and having 12 random people decide whether or not your kid is guilty of aggravated assault.” There’s no way we would bring it back because trust in one another in civil society has so evaporated. That’s what’s worrying about the decay of people’s sense of, “We live in a representative government. If the majority of us don’t want this thing and we tell our elected officials we don’t want this thing, then we’re not going to have it.” And then when you repeatedly find out that you are going to have it in spite of what you want, your ability to believe in how the system works has really deteriorated.

A lot of that sapping away of our civic strength and our capacity as citizens to deliberate with one another and hear one another out is because of the incursions of the artificial state.

I want to end by asking about a phenomenon that I see more and more, especially after AI got booed at all the college graduations. It’s amazing. Between that and the data center pushback, it seems like the industry has decided that they need a new tack. They need to communicate more. OpenAI is just running around and its executives are like, “We have a marketing problem. We just need to market it better.”

Google’s CEO was on the show , and he said, “I don’t think it’s a marketing problem. We actually have to prove that the products are useful to people.” That’s what the Microsoft CEO says. There’s a range of opinions here.

The thing that all of them have started saying over and over and over again is that the democratic process should be trusted to put the guardrails on them. The CEO of Skydio was on . He is a drone maker. He supplies drones to the military and various governments, and he will tell you it’s as much power companies as the police, but don’t worry about the police. The democratic process is going to stop the police from infringing on your rights with drones. Maybe.

I hear it with Anthropic, which just proposed a federal regulator for AI that will determine who runs the frontier models and do safety checks on them, and that’s the democratic process. I could probably make a list of every big tech executive at some point saying, “Democracy will stop us. We trust democracy.”

Your entire book is about technology fraying democracy to its very core. Is this just self-serving on their part? Now that they’ve broken it, they’re saying, “That’s the thing that will stop us,” and they know that it won’t. Or is this actually the moment for democracy to reassert itself because they’re asking for it?

Yes to both. It is self-serving on their part, and this goes pretty far back. Some years ago, OpenAI had this whole democratic inputs project where they’re going to get democratic inputs on how AI should run. This is where Claude’s constitution comes from. “We’re going to have a collective constitution. We’re going to crowdsource constitutional drafting.”

The individuals involved in the democratic inputs project for OpenAI and the collective constitutional project for Anthropic really cared about those things. They knew nothing about the political process. They knew nothing about representative democracy, nor did they care to learn. The way collective constitutionalism worked, the democratic inputs project worked was basically just turning people into data points. “We’ll do hyper-scaling. Every American will be able to point and click and vote.”

That’s not deliberative democracy. Actually, people have to be in a room together. They need to talk to one another. There needs to be a process. There need to be rules. There needs to be a parliamentarian. There need to be votes. There need to be rules about vetoing votes. There need to be rules about speech and process.

So when they talk about, “Democracy will save us,” what they mean by democracy is something that I think, for most political scientists, is unrecognizable as a democratic process. It’s a buzzword for them. It’s been convenient for them at different moments in their recent past, and it seems convenient now.

That said, that doesn’t mean this isn’t a moment that people who care about involving ordinary people in discussions around, “What do we want? What would be right? I ought to learn a little bit more about this, and I want to talk to some people, and I want to hear what my neighbors think. And I want to sit down with the town council, the neighborhood council and work it out.” That is the chief argument in my book. That can still happen. That should still happen, and I think it has to still happen.

That seems like the next step. It seems like getting the college kids together to boo is maybe the thing that will solve democracy. As a college student who booed the Iraq War, I still want to believe this, whether or not that went my way. Dr. Lepore, this has been amazing. Thank you so much.

Questions or comments? Hit us up at decoder@theverge.com. We really do read every email!

Decoder with Nilay Patel

A podcast from The Verge about big ideas and other problems.