On the campus district

I'm writing this from Mather's gardens, in what's left of a mansion built by iron ore money in 1910. Samuel Mather's fortune came from extracting resources—pulling minerals out of the ground, processing them, selling them, repeating until the wealth became incomprehensible. He was a steam locomotive running down the track. You didn't negotiate with him. You got out of the way or you got run over. He built this place to show it off. So did Rockefeller down the street. So did all of them, lined up along what was called Millionaires Row.

You've never heard of it. That's fine. Nobody has.

The mansions are mostly gone now. Torn down, repurposed, carved into administrative offices. Cleveland State University squats on the former estates. The churches the millionaires built to launder their guilt through stained glass are still here—I can see five of them from where I sit. The universities they endowed, the museums, the theaters, the whole cultural apparatus of University Circle exists because these men decided how their wealth should be remembered.

Except we don't remember them. We remember what they built. We live in the infrastructure and forget who paid for it.

Here's what I think about, sitting in these gardens: Today's wealth looks different but it's built the same way. Zuckerberg. Musk. Altman. Bezos. The new extraction isn't iron ore, it's data. Our data. Every word we type, every image we upload, every click and scroll and search. They're pulling it out of us, processing it, feeding it into models that will shape everything about how we live.

And they're running out.

That's the thing nobody wants to say clearly: they're running out of training data. The internet has been scraped clean. The books have been pirated and fed in. The images have been processed. They've extracted everything they can extract from the existing world, and the models are hitting walls. Diminishing returns. They need more, but there isn't more. Not without fundamentally changing what they're doing or how they're doing it.

So they'll pivot. They'll synthesize. They'll generate their own training data and feed it back in, an ouroboros of algorithmic output training algorithmic output. They'll scrape harder, reach deeper, violate more boundaries. They'll lobby to redefine copyright, privacy, consent—whatever stands between them and the raw material they need.

Because that's what extraction does. It doesn't stop when the easy stuff runs out. It gets more aggressive.

I think about Mather sitting here in 1920, looking out at his gardens, secure in the knowledge that his fortune had built something permanent. The mansion, the institutions, the legacy. What he didn't know—what none of them knew—is that the legacy wouldn't be about them at all. We kept the universities and forgot who endowed them. We kept the churches and forgot whose guilt built them. We live in the infrastructure of their wealth and treat it as if it simply emerged naturally from the landscape.

The tech titans are building our future right now. They're deciding what gets preserved and what gets erased, what gets encoded into the systems that will mediate our reality and what simply falls away. They're running out of material to extract, and they're going to make choices about what comes next. They're the new locomotives, and they're not slowing down just because the track is running out.

In a hundred years, will anyone remember their names? Or will we just live in the infrastructure, the algorithmic landscape they're building right now, and assume it was always there?

I'm writing from the middle of a previous extraction's legacy. On the campus district. Up the street from the theater district. Down the road from University Circle. Right here, where everyone forgot this used to be Millionaires Row.

Let's not forget this time.

---

More to come.