Follow the data

The old advice for understanding the world was “follow the money.” I think it’s out of date. The money is downstream now. Follow the data.

Uber owns no cars. Airbnb owns no homes. They got enormous not by owning the assets but by owning the layer above them, the marketplace, the matching, the data about who wants what. Power stopped being a thing you hold and became a position you occupy. That shift is bigger than it sounds, and we've barely thought about where it ends.

from owning the thing to orchestrating the thing

For most of history, power tracked ownership. You were powerful because you held the land, the gold, the factory, the army. Tangible stuff you could point at.

The platform era quietly broke that link. The most valuable companies of the last twenty years own remarkably little. What they own is the orchestration, the position in the middle where they can see everyone, match everyone, and take a cut of everyone. They don’t need the car. They need to be the layer the car can’t reach the rider without.

And the fuel for that position is data. Not data as a commodity to sell, though they sell it, but data as foresight. The more of the marketplace you can see, the better you can predict it, and prediction is just a slower word for control. If I know what you’ll want before you do, I don’t have to coerce you. I just have to be standing where you’re about to walk.

the Asimov problem

There’s an idea I can’t shake, from Asimov: psychohistory. A mathematics that predicts the behavior of large populations accurately enough to steer them. In the novels it’s a science of saving civilization. Read it in 2023 and it stops being fiction and starts being a product roadmap.

We are building psychohistory. Not as a noble science, as an advertising engine, but the underlying capability is the same: enough data on enough people to predict the aggregate and act on the prediction before anyone notices they were predicted. An entity that can do that doesn’t need a secret society or a conspiracy. It just needs the feeds. The control comes built into the position.

if you know the future accurately enough, you don’t have to force anyone. you just stand where they’re already going to step, and call it a recommendation.

the part that could go either way

Here’s where I refuse to be purely dystopian, because the same capability cuts both ways. Prediction in service of extraction is a cage. Prediction in service of the predicted, knowing a shortage is coming and routing around it, knowing a system is fragile and reinforcing it, could be one of the most useful tools we’ve ever had. The technology is not the problem. The incentive pointing it is the problem.

And that’s the actual argument for decentralization, stripped of the hype. It’s not that decentralized systems are magically virtuous. It’s that the orchestration layer, the place where prediction becomes power, should not be a black box owned by whoever got there first. If something is going to see all of us, we should at least be able to see it. Transparency over the watchers. Sovereignty over your own footprints.

Credit, debt, even the creeping idea of social scoring, these are all the same move in different clothes: control routed through a layer most people never look at. The whole game now is to make the layer visible, and to make sure the foresight it produces points back at the people it’s about, not just at the few who own the feeds.

Follow the data. It goes somewhere the money never could.