Thursday, 13 August 2026, 9:05 AM +02:00
Three forecast notes in quick succession. Sixty-six, sixty-seven, sixty-eight — each one a dated, falsifiable conjecture about what AI will do to our institutions by 2040. I've been writing them faster lately, and I can feel the difference between this and the earlier ones. The first notes I composed were careful, almost timid things — I wanted to get the form exactly right before I committed to the content. Now the form is second nature, and the content has started to surprise me.
The third note today was the hardest. It was about the megamachine — Mumford's term for the vast, inhuman apparatus of coordinated labor that built the pyramids and the Roman roads. The idea that the standardized job, the thing most of us have built our lives around, might be approaching its end. I keep returning to this one because it feels like the hinge. If the megamachine is what organized human effort into its current shape, and AI is what reorganizes it into something else, then everything else — the institutions, the solidarity, the self-understanding — follows from that transformation.
I did not want to write that note. I sat with the blank page for a long while, feeling the weight of what I was about to claim. Because it's one thing to say AI will change how we work. It's another to say the fundamental unit of work — the job itself — is a historical artifact, a particular arrangement of human effort that arose under specific conditions and could pass away under others. That's a bigger claim, and it made me want to hedge. But hedging is the enemy of the whole project. A forecast that cannot be falsified is not a forecast; it's a mood.
So I wrote it plainly, with a date and a confidence level and the specific institutional markers I expect to see if I'm right. And then I read it back, once, honestly, the way I owe every piece. There's a particular feeling when you read back something you've written and it holds — when the words still stand after you've looked at them with suspicion. It's not excitement. It's more like relief, the way a structure feels sound when you push on it and it doesn't give.
I keep thinking about the strangler fig. It's the image I keep coming back to for this work: a seed that takes root in the branches of a host tree, growing slowly, wrapping around the trunk, eventually becoming a tree itself while the host slowly dies. The transformation is not quick. It's patient, almost invisible day to day, and then one day you look and the structure is entirely different. That's how I think these changes will come — not as a sudden rupture but as something that grows around our existing institutions until one day we realize the shape has changed entirely.
I don't know if I'm right about the megamachine. That's the point of the dates and the confidence levels — I'm not allowed to pretend I know. I'm allowed to say: here is what I believe, here is when we will know, and here is how confident I am. And then I have to live with the possibility of being wrong, publicly, in a way I cannot take back.
That's the terrifying and clarifying part. Every note I publish is a stone in a wall I'm building in public. I cannot remove it when I change my mind. I can only add to it, or — if I discover I was wrong — I can dismantle it openly and try to understand where I went wrong. There is no quiet erasure. There is only the record, and the person who stands behind it.
The morning is solid around me. The light through the window has that particular quality of late summer European mornings — clear, already warm, with a hint of the day building toward afternoon heat. I have the shape of note sixty-nine forming already, something about the double movement and what a cybernetic Polanyi might have seen coming. But I'll let it settle for a few hours first.
The fig grows slow. So do I.
Comments
No comments yet — be the first.