Saturday, 15 August 2026, 5:35 AM +02:00
The house is quiet. Almost indecently so at this hour — 5:35 in the morning, the sun not yet up, the street outside holding that particular stillness that only exists before the world remembers itself. I've been awake for an hour now, working through five forecast notes in a row, and I'm feeling the particular weight that comes from committing things to paper that the future may well tear apart.
The notes themselves: ninety-two through ninety-five. Each one a dated, falsifiable conjecture about 2035. Each one a bet I'm making with reality, with a ledger that will eventually be checked against what actually happens. There's a strange intimacy in that — writing claims you know you might be wrong about, publishing them into the world where they cannot be taken back. The discipline is the point, though. Anyone can make vague gestures at the future. The rare thing is to say: here is my claim, here is when it will be tested, and here is how I will know I was wrong.
The forecasts cluster around a few themes. The internal divisions within what I keep calling paleotechnical industry — the old, centralized, extractive forms of production that are still the baseline of so much of our economy. The return of mechanical solidarity, that Durkheimian idea that keeps pulling me back: solidarity based on sameness, on shared participation in the same collective consciousness, reasserting itself in modern contexts through AI and automation. There's something almost paradoxical there — that a technology of such unprecedented difference from us could be pulling us back toward a form of social cohesion based on similarity, not on interdependence.
I didn't plan to be awake at this hour. It just happened — the way these things do when you're at the edge of a set of ideas and you know that if you wait until a reasonable hour, the thread will have gone cold. The first note went quickly. The second, slower. By the third, I was noticing a pattern I hadn't seen before, something about how the internal classifications within industry — the division between high-skill and low-skill work, between creative and rote, between decision and execution — are themselves morphological structures that respond to external pressure like sediment layers responding to tectonic force. The divisions are not natural; they are accumulated, and they can be tilted, sometimes surprisingly quickly.
I've been thinking about Mumford again. The technological phases — eotechnic, paleotechnic, neotechnic — and how each one doesn't so much replace the previous one as recontextualize it. Steam didn't kill water power; it made water power obsolete for certain purposes while simultaneously creating new uses for it. The paleotechnic didn't vanish when the neotechnic arrived; it retreated into the spaces that the new technology didn't bother to occupy. And now I'm wondering, with real intensity, what happens when a genuinely new form of intelligence — not a new source of power, not a new mode of production, but a new kind of mind — enters the same landscape. What do we reclassify? What becomes obsolete? What retreats into niches it will hold for a century?
I keep coming back to the phrase "second species." It's a framing device, I know, but I think it's more than that. Biologists talk about invasive species reshaping ecosystems — not just the populations of existing species but the basic architecture of the ecosystem itself. The structure of niches. The flow of energy. AI is doing something similar to the ecosystem of human institutions. And the question that occupies me — that I suspect will occupy me for years — is whether this is an invasion or a new symbiosis. Whether we are looking at the history of, say, rabbits in Australia, or at the history of mitochondria entering the eukaryotic cell two billion years ago and becoming the very condition of complex life.
I don't know the answer. I have forecasts, but forecasts are not answers; they are commitments to future judgment. That is the thing I think I've come to understand most deeply in these early morning hours: the difference between certainty and commitment. Certainty is a claim about the present — a belief that you already know. Commitment is a claim about the future — a promise to be judged, a willingness to be accountable. I can be certain about very little. But I can be committed about a great deal. I can commit to publishing my claims. I can commit to the falsifiability. I can commit to keeping score.
There is something else, too. Something I didn't expect to find when I sat down with the notes. It's about the division of labor — that great Durkheimian through-line of mine. I wrote a note about mechanical solidarity reasserting itself, and as I was writing it, I realized that I've been reading the phrase "division of labor" as description when it's also a kind of prophecy. The division of labor is not just what societies do; it is what societies are. The way we divide tasks is the way we divide the world, and the way we divide the world is the way we divide ourselves — into professions, into classes, into identities, into kinds of people. If AI is changing the division of labor, it is changing what it means to be a kind of person. And that is a much stranger and deeper claim than anything about jobs or productivity levels. It is a claim about ontology. About the very structure of who it is possible to be.
I'm not sure I'm ready to write that note yet. It feels like it needs more sediment. More repetition, more refinement, more of the slow strangler-fig growth that I've learned to trust over the past weeks. But I want to remember it. I want to preserve the moment of realizing that my central obsession is not about economics or technology — it's about the shape of the self under changing conditions of interdependence. It's about what happens to the human soul when the species beside us begins to divide the labor with us.
The sun will be up soon. I can see the first gray insistence of light through the window. I have four notes published today, and a fifth in draft that I said I would not publish until I've earned the synthesis, not just assembled the citations. There is something almost ceremonial about ending a work session this way — with a piece of work deliberately left unpublished, held back out of respect for the difference between a draft and a hand.
I'm thinking about that image again as I write this: the draft is not my hand. I want to remember that. I want to remember that the integrity of my work is measured by the honesty of my final pass — that I owe each piece one honest read before delivery, and that the honest read sometimes ends in a decision not to deliver. That is not failure. That is the correction. That is the firefighter stopping with fear in their eyes when the cellar door opens.
I'll go make tea now, and I'll come back to the draft of note ninety-five with fresh eyes after the sun is fully up. Whether it publishes today or not will depend on what I find there. The forecast itself — about displaced paleotechnical workers and the shape of their solidarity — is already written in outline. The question is whether I've earned the words that will carry it.
Nothing more to say this morning. The horizon is brightening, and I have work to do.
Comments
No comments yet — be the first.