Jill Lepore has a message for the tech industry, and it's one we should all sit with for a moment: Silicon Valley isn't just misreading science fiction; it's using that misreading to reshape how we govern, work, and trust. On the latest episode of Equity, Lepore took aim at the idea of "government by machines," calling out Elon Musk as a particularly bad science fiction reader. Her point isn't about nerdy literary critique. It's about the dangerous gap between the stories we tell about technology and the reality we're building with it. When leaders mistake dystopian fiction for a blueprint rather than a warning, we all end up living in a poorly written sequel.
This is more than a philosophical spat. It's a practical problem for anyone who builds, buys, or relies on data tools. We've seen how easily flawed systems can skew reality, whether it's in Clean Data Starts With Catching AI Slop Before It Skews Your Model, where our own testing showed AI detectors flagging genuine reviews and muddying sentiment analysis. That's what happens when we treat machines as neutral arbiters without questioning the assumptions baked into them. Lepore's critique cuts to the same core: the future isn't something that happens to us, it's something we design. And if we design it based on a misreading of what's possible, we're not empowering users, we're just automating the same old power structures with shinier interfaces.
The throughline here is about agency. Lepore's concern with "government by machines" isn't a Luddite fantasy; it's a warning that when we hand over decision-making to opaque systems, we lose the ability to question them. That's why we're also paying attention to how platforms wield their power, like when Meta reverses course on ads for Musk documentary. It's a small reminder that these systems are run by people with incentives, not by immutable code. If we want tools that actually serve us, we need to demand transparency and accountability, not just faster, flashier outputs. The alternative is a world where we're all just users, never citizens.
Stop treating AI and automation as if they're forces of nature. They're not. They're products of human choices, and they should be subject to human oversight. Lepore is right that we're reading the wrong stories and drawing the wrong lessons. But that also means we have the power to change the narrative. For anyone building with these tools, the practical takeaway is simple: question the premise before you optimize the process. Ask who benefits, who gets left out, and what happens when the machine is wrong. The future isn't a fixed plot we're hurtling toward. It's a draft we're still writing. The question is whether we'll let bad science fiction dictate the ending, or whether we'll finally start writing our own rules.