Glossary

The words we use, with each definition as it stands on its own page. Read from the top: each entry builds on the one before.

dark context

Dark context is what an organization knows but never wrote down — the situated context a general AI model cannot see.

The session shows the room what your AI is missing about you, and the later formats write it down and connect it.

What is dark context →

the three parts: app, model, context

Every AI setup has three parts: the app, the model and the context.

  1. The app: the AI program you use. Engineers call it the harness.
  2. The model: you rent this.
  3. The context: your specific knowledge and your setting. This one is yours.

The formats work on the third part, with the app and the model you already pay for.

How it works →

unwritten knowledge

Unwritten knowledge is everything an organization knows but never recorded: the skill, judgment, relationships, and situated context that live in people rather than in documents.

Each of the three formats works on it: the room puts it into words, and you decide what stays yours.

Unwritten knowledge →

tacit knowledge

Tacit knowledge is what people know but cannot easily put into words: the skill, judgment, and know-how learned by doing rather than by reading.

In the room, your team puts some of it into words.

Tacit knowledge →

tribal knowledge

Tribal knowledge is know-how a group shares informally but never writes down: the undocumented “how things really work here” that lives in the team rather than the docs.

Tribal knowledge just needs to be written down. In the mapping, the room writes down one thing it always knew and never wrote.

Tribal knowledge →

institutional memory

Institutional memory is the accumulated knowledge an organization carries over time: why past decisions were made, what has already been tried and abandoned, how things are actually done here.

The formats aim at a memory that stays current as people and priorities change.

Institutional memory →

the four kinds

Dark context comes in four shapes.

  • Never said (Unarticulated): "Silent knowledge. Nobody has put it into words yet. It lives in hands, habits, and the sense of how things are supposed to go."
  • Said but never written (Articulated): "Said out loud, not written. Someone could tell you if you asked. Nobody asked. Nobody wrote it down."
  • Written but lost (Lost): "Documented, but buried. No one finds it when they need it. Being written down is not the same as being reachable."
  • Fenced off (Fenced): "Documented, but gated. Only certain people can reach it. The rest of the team operates without it, and no AI sees it either."

In the mapping, you mark where your knowledge is missing in these four ways. The knowledge audit: twelve statements, two minutes, no signup.

the four memory steps

Where your company memory sits. Your AI can only see the last step.

  1. Unsaid: in a head.
  2. Verbalized: said once, in a meeting.
  3. Documented: written, out of your AI's reach.
  4. Connected: your AI reads it and acts on it.

Each of the three formats takes your memory one step up: the session says it, the mapping writes it down, the build connects it.

How it works →

the five stages

Five stages of what your AI can see, from Lights off to Kept lit.

  • Lights off: answers silent.
  • Low light: answers generic.
  • One lamp: answers someone's.
  • Corners lit: answers on point.
  • Kept lit: answers yours.

Each stage names a next step, from the session at Lights off to the monthly review at Kept lit.

Where are you? →

the context levels

Organisational knowledge in seven levels, from the tacit know-how in someone's head to context your AI can act on.

  • The seven levels: Tacit context, Articulated context, Recorded context, Findable context, Updated context, Passively connected, Actively connected.
  • Levels 1 and 2 are dark context, levels 3 and 4 gray, and levels 5 to 7 green.

Bring one process and we'll place its context on this map.

The 7 context levels →

context decay

The kinds of context inside one company decay at very different speeds. Some rot in weeks, some hold for decades.

It is the reason for the monthly review: written down does not mean up to date.

Context decay →

gray context

Context you captured and trusted, that has quietly gone out of date. Still findable, no longer true. The decay state between green context and dark context.

"Dark you never saw. Gray you stopped watching."

The monthly review looks for it: once a month, we go through what your AI got wrong with you, and fix the context behind it.

Gray context →

green context

Context that is current, findable, and queryable. Fresh enough to trust, reachable when you need it, and structured so people and machines can ask it questions. The opposite of dark and gray.

"Context that answers back."

It is what the monthly review works for: context that stays current as people and priorities change.

Green context →

conversational maps

A conversational map is a taxonomy of a field: the concepts a discipline uses when it talks about its work, collected and laid out so that everyone can see the same landscape.

In the three formats, the cards on the table give the room the parts of such a map, and the room lays out its own work.

Conversational maps →

Not sure where your team is?

Take the knowledge audit, or book a call and tell us where your dark context sits.

Knowledge audit — twelve statements, two minutes →