DATAVISM THE DATA UNDERGROUND
NETWORK / G · GHOST · FOUNDATION / G3
G G3 OPEN · LIVE NOW

INTAKE

What raw material do I need, and how do I structure it?

After this station you can turn a messy pile of links, files and notes into a structured source trail an investigation can actually run on.

G1
G2
G3
G4
G5
// SIGNAL

You have a question and a way to command the machine. Now you need material — and material arrives as a mess: links, screenshots, exports, half-remembered claims.

Most weak investigations die right here — jumping from question to conclusion with no source trail. Intake is the discipline that stops that.

// METHOD

Raw material is not evidence until it is described. Every source gets metadata: what it is, who published it, when you got it, and what it does — and does not — contain.

1List every source with an ID, type, owner and access date.
2For each: what it contains, what it does NOT contain, how far you trust it.
3Separate sources from interpretations from noise.
4If it is data-like, sketch a field map: field, meaning, example, caveat.
// TOOLING

A table is enough. AI can help classify and extract — you decide what counts.

A spreadsheet or tableAny AI assistantYour G1 question + G2 commands
// MISSION

Build the smallest real source trail for your investigation — and know its weak point.

1Collect at least 5 pieces of source material for your G1 question.
2Classify each: document, dataset, page, export, screenshot, observation.
3Put them in an intake table with reliability + caveats.
4Name your strongest and weakest source, and the smallest evidence set you actually need next.
// VERIFICATION

Your intake holds when:

→Every entry says what it cannot prove, not just what it shows.
→You can tell a source from an interpretation from noise.
→You can point at the single weakest link in the trail.
// GHOST

A pile of links is not evidence. A folder of screenshots is not evidence. A table is not evidence.

Evidence begins when material is placed in relation to a question.

Intake is where the investigation becomes real.

// BRIDGE

Now you have material and structure. Next — THE CONFIDENT LIE: how easily a fluent machine distorts the trail you just built.

Evidence starts as messy intake.