Silicon Valley Reads Dystopia as a Product Roadmap

Elon Musk names his rocket landing ships after Iain M. Banks starships. Sam Altman drops science fiction references the way other CEOs drop quarterly metrics. Harvard historian Jill Lepore has an uncomfortable question about all of it: weren’t those books supposed to be warnings?

The Warning Label Problem

Lepore’s argument is simple and it stings. Much of 20th-century science fiction was a pessimistic thought experiment — writers imagining how far technology could degrade the human condition. The authors were saying this would be bad. Silicon Valley reads the same pages and hears something else entirely: this is buildable.

Look at the pattern. Neuralink’s brain-computer interface. Mars colonization. The race toward superintelligence. Each one lifts a device straight off the page. What gets left behind is the context. In the source material, that technology usually shows up after society has already broken, or while it’s actively breaking. The gadget comes with a setting attached, and the setting is the point.

It’s the difference between reading Brave New World as a caution and reading it as a product spec for soma. One of these is a misreading. The other is a business plan.

Three Reasons the Misreading Sticks

Lepore isn’t accusing anyone of poor reading comprehension. She’s describing structure.

Engineering culture reframes problems until they’re solvable. That’s the whole discipline — decompose, scope, ship. But the questions dystopian fiction raises don’t decompose. What happens when power concentrates in a handful of hands? Can you trade human dignity for efficiency? You cannot sprint these. So they get filtered out as noise, not because anyone decided they don’t matter, but because the pipeline has no slot for them.

Science fiction makes excellent self-justification. For a founder, citing Banks or Asimov layers literary authority onto the story that you saw the future before everyone else. The warning half of the novel undercuts that story, so it quietly falls away. Nobody has to consciously discard it.

There’s no historical memory. This is Lepore’s oldest complaint, and the one she has the most receipts for. Her book These Truths and her work on Simulmatics — the 1960s company that tried to model and manipulate American voter behavior with computers, and collapsed — make the same point repeatedly. The dream of predicting and steering society with data isn’t new. It was tried. It failed. The people reaching for it again haven’t read the postmortem.

Where Democracy Enters

This is the load-bearing part of Lepore’s critique. Misreading science fiction isn’t a matter of literary taste. It becomes a theory of how societies should be run.

Notice what dystopian worlds tend to lack: legislatures. Elections. Public comment periods. Tedious negotiation. Problems get solved by a brilliant individual or an omniscient system. Adopt that as a template and democratic process starts looking like a bug — slow, inefficient, cluttered with people who don’t understand the technical details.

You can hear that sentiment in the industry’s recent output. Regulation is the enemy of innovation. Politics is a coordination problem awaiting a better protocol. Decisions belong to the people who understand the consequences. It’s the same instinct behind the AI governance pitch that runs: let us handle safety internally, we’re the ones who know how this works.

Lepore’s rebuttal is a historian’s. Democracy is slow by design, not by defect. The separation of powers, the checks, the friction — those were deliberate, built by people who had watched concentrated authority go wrong. Optimizing that friction away is not a novel idea. The 20th century ran that experiment several times. The results are on file.

The Counterargument Deserves a Hearing

Not all of this critique lands cleanly.

Technologists have a fair response: inspiration is how anything gets built. The flip phone descended from the Star Trek communicator. The tablet came out of 2001. Science fiction has always been a two-way street between imagination and engineering, and it was never uniformly bleak — Banks’s Culture novels, the ones Musk actually cites, describe a post-scarcity society that mostly works. Reading them as a warning requires some selective quoting of your own.

There’s a second problem. Humanities critique rarely arrives with implementation details. What specific procedure should govern frontier model releases? Which agency enforces it, with what authority, funded how? Reading more history does not produce a regulatory framework. Read the past is sound advice and a useless ticket in a sprint backlog.

Lepore’s camp has an answer to that too: the job of history and literature isn’t to supply solutions. It’s keeping the questions alive. Who profits. Who absorbs the cost. Why is this proceeding when nobody consented to it. Delete those questions and what remains is pure technological momentum — things happen because they can.

What to Watch For

A caveat worth stating plainly: this piece traces the through-line of Lepore’s published work and interviews rather than any single recent article, and it isn’t anchored to a live public-opinion snapshot. The argument is hers; the assembly is mine.

The frame holds up regardless. Whether the subject is AI safety commitments, the electricity that data centers are pulling off strained grids, or what recommendation algorithms do to public discourse, the same question keeps surfacing. Does building a technology grant you the authority to decide its social consequences?

So here’s the exercise. Think of a piece of science fiction you loved that has since become real. Go back and check how it ended for the people living inside it. In most of them, the technology worked exactly as designed. That was the problem.