Jill Lepore Says Tech Leaders Are Rebuilding Government — and Reading the Future All Wrong

Updated August 9, 2026 7:24 pm

Harvard historian Jill Lepore says Silicon Valley is helping replace democratic government with private systems of rule, and she warns the result is a more opaque, less accountable “artificial state.” In a new interview tied to her forthcoming book, she argues that today’s tech leaders are not just building products — they are increasingly taking over functions once reserved for elected institutions, with consequences for democracy, consent, and power.

Lepore, a Pulitzer Prize-winning historian and staff writer at The New Yorker, says the trend matters because it is not simply about innovation. It is about who makes decisions, who controls the infrastructure of daily life, and whether citizens still meaningfully govern themselves in an era shaped by platforms, algorithms, and corporate systems.

Speaking on TechCrunch’s Equity podcast, Lepore discussed her upcoming book, The Rise and Fall of the Artificial State, and made the case that the tech industry has long sold an appealing myth: that computers, networks, and now AI will naturally produce better democracy. In her view, that promise has repeatedly masked a deeper political shift toward corporate governance and away from public accountability.

What Lepore means by the “artificial state”

Lepore uses the phrase “artificial state” to describe a world in which private technology companies absorb functions traditionally performed by government. That includes communications, information distribution, identity, decision-making, and in some cases the rules that shape civic life itself.

She argues that this is not the same as criticizing technology broadly. Her complaint, she says, is about the gradual transfer of state power to corporations that were never elected, never authorized by voters, and are often governed by incentives that have little to do with democratic legitimacy.

“I’m not anti-technologist,” Lepore said in the interview, emphasizing that her objections are aimed at the political power private firms have accumulated as they take on functions once held by the state.

That distinction is important to her argument. Lepore is not claiming that computers or networks are inherently anti-democratic. Instead, she says the political danger arises when companies begin running systems that shape how people communicate, access services, make choices, and understand the world — often without public consent or meaningful oversight.

She described the transformation as partly accidental at first, driven by ideals such as speed, efficiency, and cost-cutting. But she argued that the process has become more deliberate in recent decades, as some prominent technology figures openly discuss moving beyond the nation-state or replacing its functions with machine-managed systems.

Why does Lepore see this as a threat to democracy?

Lepore sees the shift as dangerous because democracy depends on consent, accountability, and transparency, while artificial systems tend to concentrate power in private hands and make governance harder to scrutinize.

In her telling, the problem is not just that tech firms are powerful. It is that they increasingly mediate public life while presenting themselves as neutral platforms or technical utilities. When corporations decide how speech is ranked, what information surfaces, how users are categorized, or how digital systems allocate attention, they are in effect making political decisions without behaving like democratic institutions.

She also argues that the rhetoric surrounding these systems often obscures what is really happening. Efficiency can sound benign, but it can also become a cover for privatization. A promise to streamline bureaucracy can morph into a broader erosion of public authority.

In Lepore’s view, this is not just an American story, though it has been especially visible in the United States. She described the change as a gradual transformation that has spread across many governments, with the U.S. often serving as the place where the model first becomes visible.

How Silicon Valley built a political story around technology

Lepore’s argument rests on a larger history of how the tech industry has framed itself. She says the industry has long promoted an almost moral idea of progress: that better machines automatically produce better societies. That narrative, she argues, has deep roots in 19th-century ideas about evolution and technological advancement.

She sees a turning point in the modern imagination of computing in Apple’s famous 1984 Macintosh commercial, which cast the personal computer as a liberating force against the gray, monolithic world of mainframes. The ad, which echoed George Orwell’s novel 1984, presented the Macintosh as a blow against totalitarian control.

To Lepore, that was brilliant marketing, but also the beginning of a long-running confusion between consumer technology and democratic renewal. Personal computers were useful, exciting tools, she said, but there was little reason to believe they were inherently instruments of civic liberation.

The story, in her view, gets stranger as time passes. What began as product branding later hardened into a broader ideology, one in which tech leaders increasingly talk as if their companies are not merely businesses but institutions with civic obligations, constitutional responsibilities, or even quasi-sovereign authority.

From ad campaign to governing philosophy

Lepore argues that one reason this shift has been so effective is that it started with language people found optimistic and accessible. When technology is framed as liberation, it can be hard to notice when it becomes infrastructure for control.

That is especially true, she says, when the people building these systems are charismatic and confident. The language of disruption can make it seem as if old institutions are obsolete simply because new ones are faster or more popular.

But she contends that speed is not the same as legitimacy. A system may be efficient and still be anti-democratic if it concentrates authority in private hands or leaves citizens no real avenue for challenge.

Twitter, the myth of the digital town hall, and what went wrong

One of Lepore’s sharpest examples is Twitter, now X, which she says was never originally presented as a democracy machine. In her account, the platform began as a lightweight social tool, not a civic project. Its early appeal was casual, playful, even trivial.

According to Lepore, it was politicians and campaigns that transformed Twitter into a political force by adopting it as a direct channel to voters. Over time, the company itself accepted that role and began presenting the service as a kind of public square or digital town meeting.

Lepore called the idea that Twitter functioned like a real town hall “bananas,” arguing that the platform’s user base and posting patterns did not resemble a representative slice of the electorate.

Her objection is mathematical as much as philosophical. At the time when Twitter’s political influence grew, only a fraction of Americans had accounts, and even within that group, political posting was concentrated among a small, highly engaged subset of users. That meant the platform amplified the views of the most active and polarized participants rather than reflecting the public as a whole.

For Lepore, that made Twitter less a democratic mirror than a distortion machine. Elected officials who used it as a gauge of public opinion may have been reading a narrow and unrepresentative stream of online behavior as if it were the nation’s voice.

That example matters now because it foreshadows the AI debate. If social platforms could be reimagined as civic infrastructure without ever earning that status democratically, Lepore suggests, then AI products may follow the same path — only faster, more ambitiously, and with higher stakes.

How did science fiction shape today’s tech politics?

Lepore says science fiction has not only predicted the rise of machine power but also helped teach tech leaders how to imagine it. Her book traces what she calls a literary tradition of the “artificial state,” in which humans build systems that become more capable than their creators and eventually govern them.

She says this pattern stretches back to early industrial-era anxieties and becomes especially vivid once the idea of technological progress is linked with the idea of human evolution. Once machines are understood as improving over time, it becomes easy to imagine them surpassing people in intelligence, efficiency, and authority.

That is where the warning stories enter. Lepore points to works that depict machines managing human life so thoroughly that people become dependent, diminished, or outright controlled. In many of these stories, the machine does not just automate labor. It organizes society itself.

The enduring appeal of machine-run worlds

One of the examples Lepore highlights is E. M. Forster’s 1909 story “The Machine Stops.” She describes it as eerily modern because it imagines a future in which people live isolated lives in sealed rooms, receiving everything they need through a system of screens and deliveries.

The story, in her reading, is not just about technology. It is about the replacement of human agency and direct experience with a machine that mediates almost every need. The machine becomes so central that people treat it with reverence, until the system fails and their dependence is exposed.

For Lepore, this sort of fiction has become newly relevant because the world it imagines resembles many current anxieties about digital life: isolation, overmediation, dependence on screens, and the loss of public space.

She also notes that these stories often contain a paradox. They warn against total machine dependence, but they are sometimes read by tech leaders as design briefs rather than cautionary tales. That, she suggests, is where the industry’s “bad readers” come in.

Who in Silicon Valley is she talking about?

Lepore does not reduce her critique to one person, but Elon Musk features prominently in her comments as a symbol of the culture she is describing. In her view, some tech leaders appear to treat science fiction as a menu of possibilities from which to borrow imagery, authority, and political style.

She argues that this kind of reading is often shallow and selective. Instead of treating these stories as warnings, they extract the aesthetics of power, speed, hierarchy, and futurism while ignoring the moral unease that runs through much of the genre.

Her critique extends beyond Musk to executives who speak as if their companies should stand in for institutions of self-government. She pointed to examples such as Anthropic’s decision to hire a moral philosopher to help write a “constitution” for its system, and Sam Altman’s public musings about an AI president, as evidence that major AI companies increasingly adopt the language and symbols of governance.

Lepore suggested that these gestures are often theatrical, but she said they still matter because they normalize the idea that a company can borrow the appearance of constitutional authority without becoming accountable like a government.

That point also explains why she sees the AI boom as more than a technical race. The companies are not only competing to build better models; they are also competing to define what power should look like in the 21st century.

Why she says the industry is full of “bad readers”

Lepore’s phrase “bad readers” is meant to capture a broader failure of interpretation. In her view, many tech leaders read fiction, political history, and even their own products in overly literal or self-serving ways.

She argues that they often strip cautionary narratives of their warnings and then repurpose them as inspiration. That habit turns literature into branding, critique into mythology, and democratic skepticism into design confidence.

It also helps explain why so many claims about technological destiny sound inevitable. When a CEO says a new system will transform civilization, the statement may seem visionary, but Lepore sees it as part of a long tradition of overreaching futurism that mistakes wishful thinking for historical law.

She is especially skeptical of arguments that present technological transformation as unavoidable. The claim that history demands a certain outcome, she says, is often a way to shut down debate about alternatives.

What alternatives does Lepore think existed?

Lepore insists that the current internet and AI ecosystem was not predetermined. She argues that there were real political choices along the way, and those choices shaped the result.

She points to the debates around the open internet in the 1990s, including the 1996 Telecommunications Act, as a period when different regulatory and structural outcomes were possible. In her view, those decisions were influenced by a particular political climate and have been very hard to unwind since.

That historical approach matters because it reframes tech policy as a series of contingent decisions rather than inevitable progress. If things could have been different, then they can still be different now.

• Internet architecture was shaped by policy, not fate.
• Platform dominance emerged through regulation, competition, and political choices.
• AI governance is still being negotiated, which means different rules remain possible.

For Lepore, that is the critical lesson. Citizens do not have to accept every product pitch as a historical necessity. The systems that dominate public life today were built through legislation, corporate strategy, and public acquiescence — and they can be changed through those same channels.

How the AI debate echoes earlier tech hype

Lepore’s critique arrives at a moment when AI companies are making sweeping claims about what their systems will do for workers, education, medicine, governance, and productivity. Those claims often come packaged with the suggestion that resistance is both shortsighted and anti-progress.

She sees that rhetoric as part of a familiar pattern. A new tool is introduced as helpful and efficient, then becomes a site of enormous power, and then begins to acquire ideological justification after the fact.

The pattern, in her telling, is visible in social media, personal computing, and now AI. What changes is the scale of the promise. The more ambitious the technology becomes, the more likely its creators are to describe it as transformational in civic terms.

But Lepore argues that the public should be wary of such claims. Not every innovation improves democracy, and not every efficiency gain strengthens institutions. Sometimes the opposite happens.

Why this matters now

The stakes are rising because AI systems are increasingly being discussed as decision-makers, mediators, and even replacements for human authority. When that rhetoric becomes normalized, the line between civic infrastructure and private product becomes harder to see.

Lepore’s warning is not that technology should be rejected wholesale. It is that democracy cannot survive if people stop noticing when companies quietly inherit the powers of government.

That is why her book, and her interview, land as more than a cultural critique. They are an argument about sovereignty: who rules, on what basis, and with what permission.

Timeline of the ideas Lepore traces

The history Lepore outlines spans more than a century, but several moments stand out as especially important to her argument.

What should readers take away?

Lepore’s central message is that technology is not destiny. The tools, platforms, and AI systems shaping modern life are built within political systems, and they can either reinforce or weaken democracy depending on who controls them and how they are governed.

She is warning that Silicon Valley has begun to imagine itself not as part of a democratic ecosystem, but as a substitute for it. That is what makes her book and her comments feel so timely: they challenge the assumption that the future will be made safer or freer simply because it is more automated.

Her broader point is more unsettling. If private corporations continue to absorb the roles of government while presenting themselves as neutral builders of progress, then the public may gradually lose the habit of expecting democracy to do the work of self-rule.

For Lepore, that is the real danger of the artificial state. It does not arrive all at once. It accumulates through design choices, business incentives, policy gaps, and seductive stories about the future.

And once that happens, she argues, the hardest thing to recover is not a technology. It is the idea that citizens are supposed to govern themselves.