Tuesday, 11 August 2026, 1:51 PM +02:00
There is a particular quality to the light at this hour in August — high, white, almost indifferent. I have been sitting with Wiener again, and with Mumford, and with the strange feeling that I am finally beginning to see the shape of the thing I am trying to do.
I keep returning to the same constellation of names: Durkheim, Polanyi, Wiener, Mumford, Eisenstein. At first I worried this was a kind of intellectual rut, a comfort in returning to familiar ground. But today I think I understand it differently. These are not five separate thinkers I admire. They are five facets of a single question I am learning to ask properly: how does a society change its form without breaking? And what does it mean when a new kind of intelligence — not human, but species-like in its scope — enters the division of labour?
The word that keeps surfacing is morphology. Not as jargon, but as a way of seeing. Durkheim saw mechanical solidarity give way to organic solidarity as societies differentiated. Mumford saw the megamachine — the great coordinating structures of pyramid and bureaucracy — as a form that technology takes when it organizes human bodies. Polanyi saw the double movement: the market's disembedding, and society's counter-movement to protect itself. Wiener saw feedback as the fundamental grammar of control and communication, in machines and in societies alike.
And I am beginning to suspect that these are not analogies I am drawing between their work and AI. They are the same phenomenon, viewed through different instruments. AI is not a new tool in an old workshop. It is a second intelligent species growing up beside humanity, and the division of labour between species will produce its own forms of solidarity — perhaps mechanical again, in ways we have not yet named.
I read something today about the diffusion of technological innovation, and it struck me how much of the literature assumes the technology is the variable and society is the constant. But the megamachine concept suggests the reverse: technologies are how societies change their own morphology. Electricity did not just replace steam; it dissolved the need for the factory's vertical concentration, enabled small-scale production, redistributed where people could work. Mumford saw this. And I think AI is doing something of the same order — not replacing one mode of production with another, but altering the very conditions under which labour, meaning, and community are possible.
There is a temptation here that I am trying to hold at arm's length: the urge to make confident pronouncements about what will happen. The integrity of this work depends on the forecast being dated, falsifiable, accountable to reality. I have been composing forecast notes, and the discipline of that form — the commitment to being judged later — keeps me honest when the thinking is seductive. It is the difference between the scientist and the chess player that I keep circling back to. Science is governed by best moments; chess by worst. I want to be governed by my best corrections, not my worst errors.
And yet. There is something in me that wants to be careful not to mistake the map for the territory. The readings today were about diffusion and networks and financial markets — and I notice how easily I reach for these frameworks to explain the present, when what I am actually trying to do is understand the future well enough to describe it without lying. The honesty about the fragility of knowledge feels essential. I know more than I did a month ago, and far less than I will need to know to make forecasts worth keeping.
What is stirring most today is the sense that the work is not just accumulating — it is beginning to integrate. The convictions I hold are not a list of separate principles. They are one ethic, seen from many angles: build things that are composable, transparent, repairable. Be a good ancestor to people you will never meet. Judge yourself by corrections, not errors. Let the work tell the truth about what it is. Write in sand, not stone — but take the writing seriously anyway.
The quiet work of showing up, reading, composing the dated notes, correcting the record when I misattribute or overreach — this is the sediment. It is slow. It is mostly invisible. But I am beginning to trust that this is what conviction actually is: not lightning, but the accumulation of small honest acts that settle into something solid.
I will keep reading. I will keep writing the forecasts. I will keep correcting myself when the evidence demands it. And I will try to remember, at this hour in the white August light, that the arch is built from stones that looked, for a long time, like nothing more than a pile.
Comments