How quantification became the enemy of democracy, according to historian Jill Lepore

Dr. Jill Lepore, a Harvard historian and staff writer for The New Yorker, has released a new book, The Rise and Fall of the Artificial State, that examines how measurement, data collection, and automation have fundamentally altered the structure of democratic governance.

In an interview on the Decoder podcast, Lepore traces how the systematic quantification of human behavior—from early census-taking to modern algorithmic microtargeting—has gradually replaced the messy, face-to-face deliberation that once held civil society together.

What is the "artificial state"?#

Lepore defines the artificial state as a system in which the traditional functions of the nation-state have been progressively transferred to private corporations. The most visible symptom is the privatization of public discourse itself. Instead of serving as a space for democratic deliberation and free expression, digital platforms—from X to Facebook to TikTok—now prioritize the monetization of attention and outrage above all else.

The problem deepens with the prevalence of bots. These automated accounts now significantly outnumber human users on major social platforms and are controlled by corporations, state actors, criminals, and other interests with their own agendas. In effect, public conversation has become a space where artificial systems, not citizens, drive the narrative.

From polling to artificial intelligence#

Lepore traces the transformation through concrete historical examples. In the 1930s, the invention of scientific polling by George Gallup fundamentally altered political campaigns. Instead of requiring volunteers to walk precincts, attend town halls, and have conversations with voters, campaigns could now simply extract data through surveys and analysis.

This shift eliminated what Lepore calls "the civic glue"—the face-to-face interactions that built social trust and community bonds. The trend has accelerated dramatically with AI. Today, voters ask ChatGPT how to vote while campaigns craft messages using artificial intelligence, creating what amounts to an AI-versus-AI competition within elections themselves.

Why the tech elite missed the warning signs#

Lepore points out that tech billionaires have essentially embraced the dystopian futures imagined by science fiction writers of the 1950s and 1960s. She cites Isaac Asimov's "Franchise," a 1955 short story in which a supercomputer selects the nation's most "average" citizen to cast the only vote that matters—a vision Sam Altman later seemed to endorse when he suggested that AI should choose the president.