The people building our AI future are reading the wrong books, and worse, they're reading them wrong.
The Summary
- Harvard historian Jill Lepore argues Silicon Valley leaders are "bad science fiction readers" who mistake dystopian warnings for instruction manuals
- Her upcoming book "The Rise and Fall of the Artificial State" connects tech's soaring governmental language (Twitter's "town hall," Claude's "constitution") to a pattern of reimagining companies as nation-states
- Tech founders read Asimov, Dick, and Gibson as blueprints rather than cautionary tales, fundamentally misunderstanding the genre's purpose
The Signal
Jill Lepore, a Pulitzer Prize finalist and Harvard historian, has spent her career studying how Americans govern themselves. Now she's turned that lens on Silicon Valley and found something troubling: the tech industry's leaders treat science fiction like a product roadmap when they should be reading it as a warning label.
Her argument centers on what she calls "government by machines." Tech companies don't just build products anymore. They build systems that function like governments, complete with their own constitutions, town halls, and judicial processes. Twitter marketed itself as a "town hall in your pocket." Anthropic gave Claude a constitution. These aren't metaphors. They're declarations of intent.
"The tech industry is led by bad readers who are undermining democracy."
The science fiction problem runs deeper than bad literary analysis. When Elon Musk or Sam Altman reference Isaac Asimov's robot laws or Philip K. Dick's precogs, they're not engaging with the stories' actual themes of control, autonomy, and unintended consequences. They're cherry-picking cool ideas and stripping out the context. Asimov wrote "I, Robot" to explore the limits of programmed morality. Silicon Valley read it as a to-do list for AGI governance.
This matters because these misreadings shape real products that billions of people use. If you think "1984" is an operations manual rather than a warning, you build different surveillance tools. If you read "Neuromancer" and see aspiration instead of alienation, you build a different metaverse.
Key points from Lepore's thesis:
- Tech companies increasingly describe themselves using governmental language and structures
- Founders treat dystopian sci-fi as aspirational rather than cautionary
- This literary blindness translates directly into product decisions that affect democratic institutions
Lepore's book, "The Rise and Fall of the Artificial State," arrives at a moment when AI companies are making explicit governance claims. OpenAI talks about "alignment" as if it's constitutional law. Meta's Oversight Board mimics a supreme court. These aren't just PR moves. They're assertions that private companies should perform governmental functions without democratic accountability.
The "bad reader" critique cuts to motive. It's one thing to say tech founders are building poorly. It's another to say they're building badly because they fundamentally misunderstand the cultural texts they're drawing from. Science fiction was written to make us think critically about technology's effects on society. Silicon Valley reads it as a feature request list.
The Implication
If Lepore is right, the fix isn't better regulation or more thoughtful AI policy. It's better reading comprehension at the top of the tech stack. The next generation of founders needs to understand that when Ursula K. Le Guin or Octavia Butler wrote about future societies, they weren't pitching products. They were asking questions about power, control, and what we lose when we automate human judgment.
For anyone building in the agent economy or tokenized ownership space, this is a mirror check. Are you building systems that enhance human agency or replace it? Are you reading your influences as warnings or as blueprints? The difference determines whether Web4 serves people or governs them.