Why Cassandra Always Sounds Paranoid

I was listening to Sam Harris this morning, his most recent Making Sense episode with Gary Kasparov, when a single sentence stopped me cold. Sam was describing a particular kind of person, someone with a deep knowledge of history and a finely tuned instinct for the early signs of democratic erosion, and he observed that such people “can wind up sounding fairly paranoid to normal people who aren’t tracking these things so closely.”

I had to pause the podcast and sit with that for a while. Because he was describing me. And maybe you, if you’ve found your way to this blog.

I’ve been creating notes for some time now about what I see as converging threats to American democracy and economic stability, about artificial intelligence displacing workers at a scale we are not remotely prepared for, about the concentration of power in both government and technology becoming self reinforcing and increasingly difficult to reverse, about democratic norms eroding in ways that feel incremental until suddenly they don’t. And the response I get from people I know and love, smart and decent people, is often some version of a patient smile. The subtext is clear enough. There he goes again.

So this morning I decided I actually wanted to understand this gap, not just feel frustrated by it, but dig into why it exists.

The first thing I came to is that democratic erosion and systemic risk are fundamentally invisible to normal human threat detection. Our brains were shaped over hundreds of thousands of years to respond to immediate, concrete, physical danger. A predator. A storm. Someone raising a fist. The things I write about operate at a completely different level, the level of institutional decay, of shifting incentive structures, of historical trajectory. Perceiving those threats requires a kind of second order awareness that most people haven’t developed, and honestly don’t need for navigating daily life.

And daily life is the key thing here. The lived experience of most Americans right now is that the grocery store is open, the kids got to school, the paycheck cleared. The system’s lights are still on. When someone like me shows up telling them the lights are on a trajectory toward going out, they look around, see the lights, and file me under unreliable narrator. They aren’t stupid. They’re just using the evidence directly in front of them, which is the evidence that everything still works.

There’s a social mechanism layered on top of the cognitive one. Once someone has categorized a concern as paranoid, they can dismiss everything inside that category without engaging with any of its specifics. It’s extraordinarily cheap cognition. And it protects people from something genuinely uncomfortable, because if you really took seriously the implications of what researchers and historians are saying about AI driven unemployment or the structural fragility of democratic institutions, you would feel obligated to act, or at minimum to feel genuinely frightened. Dismissal is much easier to live with than either of those alternatives.

The people who tend not to dismiss it are those who’ve built pattern recognition from specific kinds of experience. Historians who’ve watched this film before. People who grew up under authoritarian governments, which is precisely why Kasparov belongs in this conversation. Economists who model systemic risk for a living. People who’ve spent years inside the technology industry watching AI develop and understanding what it actually can and cannot do. These people are triangulating across multiple streams of information simultaneously, and the picture that emerges from that triangulation is alarming in ways that a single data point never would be.

What I keep returning to is the temporal problem. The things that make a concern sound paranoid today are often exactly the things that will make it seem obvious in retrospect. The window between “that sounds alarmist” and “why didn’t anyone do something” can be surprisingly short, and the tragedy is that meaningful intervention is almost always only possible during the period when most people still think the alarm is false.

Kasparov knows this from personal experience. He watched the erosion of Russian institutions happen in a way that felt gradual and then sudden, and he has spent years since trying to get Western audiences to understand what the early stages look like from the inside. He sounds alarmist to a lot of people too.

I’m not comparing myself to Kasparov. His credentials on this subject are built from a life I haven’t lived. But I recognize the dynamic he operates inside, because I operate inside a version of it every time I publish something here and watch it land with a polite shrug.

Cassandra wasn’t wrong. She was just right before it was convenient to believe her.