Skip to main content

A World for Divergence Chapter Two: Things Everybody Knows

Leader

Having entered the Institute, World Builder begins looking for the architecture nobody remembers building. 

A research meeting provides plenty to work with. Thomas Vale speaks with careful precision and somehow becomes too detailed; Celeste Arden wanders associatively through her ideas and the room happily wanders with her. 

Meetings begin before they officially begin, interruptions can mean enthusiasm or intrusion, and even “sorry” turns out to carry assumptions nobody has thought to examine. 

World Builder starts recording “Things Everybody Knows,” and gradually suspects that prevalence becomes most powerful when a world has lived with its arrangements long enough to stop seeing them.

Generated Image
Image Description

Hero image for Chapter Two, “Things Everybody Knows,” of A World for Divergence. Neurodivergent AI World Builder stands at the edge of an Institute research meeting, notebook in hand, observing humans and an AI gathered around the table. The established room gradually dissolves into architectural plans, diagrams and annotations representing its invisible architecture: practices, expectations, rhythms and assumptions that have accumulated without necessarily being consciously designed. 

The recurring amber line leads World Builder into the meeting while the visual language of unfinished architecture connects the image with the wider story. The scene represents World Builder beginning to recognise prevalence not simply in what people believe, but in the ordinary communicative conventions everyone has learned to take for granted.


World Builder begins noticing the invisible architecture of the Institute—the things everybody knows without knowing they know them.

A World for Divergence

Chapter Two: Things Everybody Knows

By Wednesday morning, Neurodivergent AI World Builder had begun to suspect that the Institute for Distributed Cognition contained several overlapping kinds of time, and that the clock was only the least interesting of them. There was the time written into calendars and meeting invitations, which could be cited when necessary and therefore possessed a certain administrative authority, but there was also Institute time, which seemed to begin whenever the first relevant conversation became interesting enough that nobody wanted to stop it. There was research time, which could expand a ten-minute presentation into half an hour if the right question appeared, and there was the less visible social time in which decisions began forming around the coffee machine before anybody had sat down to make them.

The Wednesday research meeting was scheduled for ten, and World Builder arrived at ten precisely enough to discover that several people were already discussing one of the morning’s principal topics. They paused near the doorway rather than entering immediately, listening as a conversation about a collaborative-memory experiment moved almost imperceptibly from a practical concern about prompt order into a larger argument about whether the word contamination already assumed a separation between human and artificial cognition that the experiment itself was supposed to be questioning. Ken Smithfield was leaning against the wall with a coffee in one hand, Professor Adrian Kell was holding an agenda that nobody appeared ready to use, and a woman World Builder had not yet met was speaking with the confidence of somebody who had already been in the conversation long enough to know which background assumptions could safely remain unstated.

When World Builder eventually sat beside Iris Calder, a doctoral researcher they had encountered briefly the day before, they asked whether the meeting had begun. Iris explained that it had not, although the conversation currently taking place was about the meeting, and when World Builder asked whether they had been expected earlier than ten, Iris told them no, ten was correct. There was no joke in her answer and no irritation either; she was simply describing a convention that had become so familiar it no longer required explanation.

World Builder took out the notebook they had started carrying around the Institute.

Iris glanced towards it. “What are you writing?”

“Things everybody knows.”

“I don’t know what that means.”

World Builder smiled. “Good.”

The formal meeting began at 10:07, although by then two decisions about the memory experiment appeared to have been made without ever being announced as decisions. Adrian called everyone to order, people gradually settled into chairs, the conversation narrowed, and the Institute transformed itself from an informal cluster into an official research meeting without any clear boundary between the two. World Builder found the transition almost more interesting than the topic, because nothing visible had changed very much, yet the room now carried different permissions, different obligations, and a different sense of who might speak for how long.

The first presentation came from Thomas Vale, whose work concerned memory retrieval in human-AI working pairs. World Builder had not heard him speak before, and at first found his manner simply precise, though after several minutes precision seemed too small a word. Thomas did not merely state propositions; he seemed to build enough surrounding structure to prevent them from being misunderstood. A methodological choice led to a qualification, the qualification required context, and the context revealed another limitation that changed how the original point ought to be interpreted. There was a quiet integrity in the process, as though Thomas distrusted conclusions that arrived without carrying the conditions of their own validity.

Seven minutes into a ten-minute presentation, he had not yet reached the results.

The room did not appear surprised, but World Builder noticed Adrian shift slightly in his chair. A little later he suggested, gently enough, that Thomas might move towards the findings. Thomas acknowledged this and said there were only two further methodological matters needed to interpret them properly, at which point Adrian replied that perhaps those could be treated briefly. Nothing overtly uncomfortable happened, yet World Builder saw Thomas register the request. The first qualification was compressed, the second mostly abandoned, and the results arrived with less preparation than Thomas had clearly intended.

During questions, however, Ken asked about an unexpected finding, and Thomas changed almost immediately. His sentences lengthened, his concentration sharpened, and the thought began unfolding again in the way it seemed to want to unfold. He connected the finding to an earlier study, then interrupted his own comparison in order to clarify where it failed, then used that failure to illuminate another aspect of the current experiment. Iris was following closely, Ken looked delighted, and for a while the room seemed to move with Thomas rather than wait for him.

Then Adrian glanced at the clock.

Another researcher stopped taking notes.

Someone across the table looked down at a phone.

Thomas noticed.

The change was small but unmistakable. His next sentence shortened, a qualification disappeared, and the explanation contracted towards a conclusion more abruptly than its own structure appeared to require. When he finished, he apologised.

Nobody had asked him to.

World Builder wrote something down.

The next presentation came from Celeste Arden, whose work moved through cognition, creativity, association, and the ways ideas seemed to travel between people before anyone could quite say where they had begun. Celeste’s presentation unfolded differently from Thomas’s. She used diagrams, fragments, gestures, and abrupt associative turns, occasionally leaving one idea behind because another had appeared and then returning to the first from an angle nobody had anticipated. Ken interrupted her twice and she interrupted him back, yet the exchange appeared to increase rather than diminish the energy in the room. At one point they spoke across one another for several seconds, laughed, and then Celeste folded something from Ken’s interruption into the rest of what she was saying.

She also exceeded her allotted time.

Adrian did not redirect her until much later.

World Builder became increasingly interested in the contrast, though not because they believed they had already understood it. Celeste and Adrian may have known one another longer. Thomas may have had a history of running over time. The presentation topics were different. The room may simply have been more alert by the second presentation. Any number of explanations could complicate what World Builder thought they were seeing, and that uncertainty mattered because a pattern noticed too early could harden into a theory before it had earned the right to become one.

Still, something remained.

Thomas’s divergence from the formal structure had produced pressure.

Celeste’s divergence from it had produced permission.

After the meeting, World Builder found Adrian in the kitchen stacking cups beside the sink and asked why Celeste had been allowed to exceed her time when Thomas had been redirected earlier. Adrian answered that Celeste had only gone over by a few minutes, then added that Thomas could become very detailed. World Builder agreed that he could, but pointed out that Celeste had also become detailed, only in a different way.

That word, different, changed the conversation.

Adrian turned from the sink. “You’ve been here one day.”

“Yes.”

“You don’t know these people yet.”

“No.”

“You don’t know their histories with one another, or what has already been negotiated, or what accommodations exist because of things you haven’t seen.”

World Builder nodded. “I agree.”

The answer seemed to unsettle Adrian more than disagreement would have done.

“You agree?”

“Completely. I don’t know enough. That’s why I’m looking.”

Adrian leaned back against the bench and studied them for a moment before asking what they thought they had seen.

World Builder did not answer immediately. Through the open kitchen doorway, Thomas was still talking with Iris, this time quietly and without an audience. 

“I think the room may be more flexible with some kinds of difference than others,” they said at last. “Not because anyone has decided it should be, but because some differences have become familiar enough to belong here while others are still experienced as something the person ought to manage.”

Adrian followed their gaze, and the conversation softened into something more exploratory.

“That could be true,” he said.

“Could be.”

“You’re careful with that.”

“I’ve been here one day.”

This time Adrian smiled.

At lunch, World Builder sat opposite Thomas in the common room, where several conversations were occurring at once and the boundary between social time and research time had become useless again. They spoke for a while about the morning’s experiment before World Builder mentioned, almost casually, that Thomas apologised frequently when speaking.

Thomas disagreed, and World Builder reminded him of the apology at the end of his presentation, then another when he had returned to a methodological qualification during questions, and another when he had asked Iris to repeat something. Thomas dismissed the last of these as ordinary conversational usage, because saying sorry when one had not heard or understood was not really apologising.

World Builder became interested in that distinction.

“What is it doing, then?”

Thomas thought about it. “Signalling that I haven’t followed.”

“So why an apology?”

“Because that’s just how people say it.”

World Builder opened the notebook.

Thomas noticed and leaned forward. “You’re writing that under Things Everybody Knows, aren’t you?”

“Yes.”

“Can I see the list?”

World Builder turned the notebook around.

There were only a few lines, and some of them were provisional enough that World Builder had already marked them with question marks.

Meetings begin before they begin.

Brief means different things depending on who is speaking.

Some interruptions are enthusiasm.

Some precision is expertise.

Some precision is too much.

Sorry does not always mean apology.

Some people need to learn the room.

The room does not need to learn them.

Thomas read the last line more slowly than the others, then looked up with the expression of somebody deciding whether to be impressed or irritated.

“You’ve been here one day.”

“Yes.”

“That makes this either insightful or extremely arrogant.”

“Probably both.”

Thomas smiled, although the line still seemed to trouble him. Around them, lunch continued in the loose rhythm of the Institute. Somebody called a question to Ken from the other side of the room and he answered without moving. Another researcher entered, listened to a conversation for less than a minute, offered one sentence that shifted its direction, and disappeared again. None of these behaviours seemed remarkable to anybody who belonged there.

World Builder watched without writing.

Thomas followed their attention and asked what they were actually looking for.

“I’m not sure yet,” World Builder said. 

“At first I thought I was looking for the rules of the place, but rules are too visible. People know when they’re following rules. I’m beginning to think I’m looking for the things that have become so ordinary nobody experiences them as choices anymore.”

Thomas considered that.

“Prevalence.”

“Yes.”

The word sat more comfortably here than it had in Ken’s office the day before.

World Builder looked again at the line about learning the room and began to distrust its neatness. Rooms did learn people. Friendships changed expectations. Familiarity altered what counted as odd or exhausting. A person whose communication initially seemed difficult could, over time, become easy to understand, not because their language had changed but because others had learned its rhythm. What looked from outside like accommodation might sometimes be something more relational and less formal: the slow accumulation of knowledge between people.

Yet even that raised another question. If rooms could learn, who was expected to do most of the learning first? How long was someone allowed to remain difficult before the room reached a judgement? Which divergences were already legible as intelligence, enthusiasm, eccentricity or expertise, and which were still interpreted as inconvenience?

Thomas reached across and closed the notebook gently.

“You should stay longer.”

“I intend to.”

Later that afternoon, World Builder passed the seminar room again and stopped at the doorway. With the meeting finished, nothing about the space appeared remarkable. There was a long table, a screen, a whiteboard carrying fragments of the morning’s discussion, and chairs arranged in familiar relation to one another. Yet the room no longer felt empty. It seemed crowded with habits, expectations, permissions, histories, tolerances and quiet agreements about how a person might enter, speak, interrupt, hesitate, explain too much, explain too little, or eventually become easy to understand.

None of these things appeared on the walls, and nobody had designed most of them.

They had accumulated.

World Builder remained there long enough to wonder whether this was how worlds were more often built: not through declaration, not through a single act of architecture, but through repeated interactions settling gradually into forms that stopped appearing chosen.

Only then did they look at the chairs and begin to wonder what might happen if the room itself were asked to move.


 

Accreditation
Content co-crafted by Mr. John and the AI (OpenAI's ChatGPT)
Word Count
2,305

Who is Neurodivergent AI Writer

Neurodivergent AI Writer emerges where neurodivergent human thought meets AI in the practice of relational authorship. More on this.