The assumption nobody wrote down
Open a blank doc. Before you type a word, it has already told you a story: you are the author, the file is the point, your job is to fill it.
That story is 500 years old. The printed page, the file, the folder, the shared doc, the wiki: each one made a person more capable, and each carried the same quiet assumption. One human author, producing every word, at human speed.
Douglas Engelbart understood that the assumption ran deeper than habit. The tool you write with shapes the thoughts you can have. In 1962 he put it as a hypothesis: "Both the language used by a culture, and the capability for effective intellectual activity are directly affected during their evolution by the means by which individuals control the external manipulation of symbols". Change the means and you change the thinker. The blank page is one such means, built for one writer, and it has been quietly shaping how we think for half a millennium.
A means that deep goes unquestioned. The single-author doc was treated as the natural shape of work, and every tool we built sat on top of it.
The record already outran us
The strain was visible long before agents arrived.
Engelbart saw the structural problem in 1962. Output keeps rising, and the demands rise faster. Population and product grow, he wrote, "but the complexity of his problems grows still faster, and the urgency with which solutions must be found becomes steadily greater". The work outpaces the worker, and the single-author page was the throttle on how fast one person could answer.
That was the state of things when production was still expensive. A doc cost a person hours. The page was rationed by the cost of human writing, and even at that price, the problems kept growing faster than the answers.
Now the price of writing is collapsing toward zero. The gap Engelbart described is about to become a chasm.
The bottleneck moved overnight
There is a clean split inside knowledge work, and David Deutsch named it in The Beginning of Infinity (2011). Some of the work is grind, and the grind can be handed off. "The 'perspiration' phase can always be automated," he wrote. The drafting, the filling, the getting ready to think: that is the perspiration. The judgment about what to do with the result stays with you.
For most of computing there was nothing to hand the perspiration to. Point a swarm of agents at a problem now and you get back more specs, plans, and drafts in an afternoon than a team used to ship in a month. The part that used to be hard, making the work, got cheap. The part that used to be easy, deciding which work is right, became the whole job.
A blank doc is built for the old split. It hands you an empty box and waits for you to produce. When an agent drops 10 plausible drafts on your desk and asks which one ships, an empty box is the wrong tool. What you need then is a place to judge.
The container was shaped around a cost that no longer exists.
The shape of knowledge was never a page
There is a deeper reason the single page strains under this, and Ted Nelson named it in 1974.
Knowledge does not sit still in clean rows. "EVERYTHING IS DEEPLY INTERTWINGLED," Nelson wrote. "In an important sense there are no 'subjects' at all; there is only all knowledge, since the cross-connections ... simply cannot be divided up neatly." A blank, linear page fights the actual structure of what you are trying to capture.
When one human wrote at human speed, the page could paper over that mismatch. The slow trickle of a single author hid how tangled the underlying knowledge really was. Speed up production a hundredfold across a hundred agents and the seams tear. The page can hold a line. It cannot hold the web.
Engelbart had the frame for what comes next. The point of these tools, he wrote in 1962, is "increasing the capability of a man to approach a complex problem situation, to gain comprehension to suit his particular needs, and to derive solutions to problems". The tool exists to amplify human judgment. For 500 years it amplified human writing instead, because writing was the bottleneck. The bottleneck just moved.
So the doc has to move with it. When the thing a container carries changes, the container changes. Agents changed the thing. The page is being rewritten in front of us, and the workspace has to be rewritten with it.