Jill Lepore on the ‘Artificial State’ and why Silicon Valley’s leaders are bad sci-fi readers

The intersection of political science and computer science has never been more fraught. As Silicon Valley pivots from social media dominance to the pursuit of Artificial General Intelligence (AGI), a new paradigm of corporate governance is emerging. Pulitzer Prize-winning historian Jill Lepore, in her forthcoming work *The Rise and Fall of the Artificial State*, identifies a concerning trend: the tech elite’s desire to treat their platforms and AI models not merely as products, but as sovereign entities. This movement represents a shift from the digital "town square" to a digital "leviathan," where corporations bypass traditional democratic institutions to install their own constitutional frameworks.

Lepore’s critique finds its roots in the specific history of Silicon Valley’s relationship with literature and statecraft. For decades, the industry has operated under a brand of techno-optimism that mirrors the "Great Men" theory of history, suggesting that a few visionary founders can—and should—reorder society. The background here is one of institutional erosion. As public trust in traditional government wanes, tech giants have stepped into the vacuum, framing their algorithms as more efficient, objective arbiters of truth than the messy, slow-moving processes of representative democracy. Lepore argues that this is not an accidental byproduct of innovation but a deliberate ideological project.

The technical mechanics of this shift are most visible in the current AI arms race. A prime example is Anthropic’s "Constitutional AI," a training method that provides large language models with a written set of principles to guide their outputs. While marketed as a safety measure to ensure ethical behavior, Lepore views these developments through a more cynical lens. By embedding a "constitution" within code, companies are effectively privatizing the rule of law. When a machine’s behavior is governed by a corporate manifesto rather than public legislation, the company ceases to be a vendor and begins to function as an "Artificial State," exercising power without the accountability of the ballot box.

This ambition is fueled by what Lepore describes as a profound misreading of science fiction. The tech billionaire class, she suggests, treats dystopian warnings like *Snow Crash* or *Neuromancer* as blueprints rather than cautionary tales. Where authors envisioned the "Metaverse" or the "Matrix" as dehumanizing traps, current industry leaders see them as inevitable markets to be conquered. This literary illiteracy has resulted in a generation of leaders who lack the historical and philosophical guardrails necessary to wield the power they are building. They are, in essence, trying to build a future based on the very scenarios sci-fi warned would lead to societal collapse.

The implications for the global market and regulatory landscape are immense. If tech companies successfully position themselves as "Artificial States," they create a jurisdictional challenge that traditional nations are ill-equipped to handle. We are already seeing this in the way AI companies negotiate directly with foreign heads of state, bypassing the usual diplomatic channels. Competitively, the race is no longer just about who has the fastest chip or the largest dataset; it is about who can create the most comprehensive digital ecosystem—one that manages the economic, social, and political lives of its users with totalizing efficiency.

Looking ahead, the tension between the Artificial State and the Westphalian State will likely reach a breaking point. We must watch how the European Union’s AI Act and potential U.S. federal regulations attempt to reclaim sovereignty over these digital territories. Furthermore, the public’s willingness to trade democratic agency for the convenience of automated governance will be the ultimate test of Lepore’s thesis. As Silicon Valley continues to iterate on its "constitutions," the rest of the world must decide whether it wants to be citizens of a republic or users of a corporate state. The danger is not that AI will become sentient, but that it will become the default infrastructure of a world where public accountability has been rendered obsolete by design.