First Principles

1. What media is

Media is a controlled change in audience state over time.

AUDIENCE STATE(t)
= knowledge
+ belief
+ emotion
+ attention
+ expectation
+ intention

A production is successful when the sequence of signals causes the intended state changes while preserving truth, context, continuity, accessibility, and technical integrity.

2. Fundamental objects

Source

An immutable recording, generated asset, document, image, audio file, data set, or live input.

Source = signal + origin + rights + technical metadata + provenance

Time range

A bounded interval in a source or output.

Range = source_id + start_time + end_time + handles

Transcript unit

Speech aligned to time.

TranscriptUnit
= words + speaker + timestamps + confidence + context + semantic tags

Proposition

The smallest unit that can be evaluated as a statement, question, instruction, or implication.

Claim

A proposition presented as true, probable, remembered, predicted, or believed.

Beat

The smallest meaningful change in audience knowledge, emotion, expectation, or intention.

Beat
= input_state
+ stimulus
+ change
+ output_state
+ story_function

Shot

A continuous visual perspective over an interval.

Shot
= subject
+ action
+ world
+ camera state
+ lighting state
+ visual treatment
+ time behavior

Sound event

A bounded auditory object: dialogue, voice-over, music, ambience, room tone, sound effect, Foley, or silence.

Graphic

A symbolic visual transformation of information.

Graphic
= claim
+ relationship
+ data
+ encoding
+ hierarchy
+ motion
+ source

Cut

A boundary that changes at least one of the following:

viewpoint | time | space | information | emotion | rhythm | sound perspective

Scene

A group of beats governed by a shared local goal, question, conflict, place, time, or interaction.

Sequence

An ordered set of scenes that resolves or transforms a larger question.

Episode

A deliverable that fulfills a recurring audience promise.

Version

A derived output that preserves semantic intent while changing format, duration, framing, language, pacing, or platform profile.

3. The editorial equation

EDIT
= SELECT
+ BOUND
+ ORDER
+ DURATION
+ RELATE
+ EMPHASIZE
+ TRANSFORM
+ MIX
+ DELIVER

Where:

• Select chooses source intervals.
• Bound determines in/out points and handles.
• Order creates sequence and causality.
• Duration controls attention and rhythm.
• Relate defines how adjacent objects interact.
• Emphasize directs audience focus.
• Transform changes picture, sound, text, or speed.
• Mix balances simultaneous signals.
• Deliver maps the semantic edit into a technical profile.

4. The story equation

STORY
= STATE A
+ DESIRE OR QUESTION
+ FORCE OR OBSTACLE
+ ATTEMPT OR INVESTIGATION
+ CHANGE
+ STATE B
+ MEANING

A sequence of facts is not automatically a story. A story requires perceived change and a relationship among events.

5. The context equation

MEANING
= content
+ speaker
+ referents
+ preceding causes
+ following consequences
+ situation
+ audience assumptions
+ evidence status
+ tone

A clip is context-complete only when an intended viewer can correctly interpret its core meaning without relying on omitted information that materially changes the statement.

6. The camera equation

CAMERA STATE
= shot scale
+ azimuth
+ elevation
+ height
+ roll
+ lens/FOV
+ distance
+ movement
+ focus
+ composition
+ frame rate
+ exposure
+ continuity

7. The lighting equation

LIGHTING STATE
= purpose
+ source roles
+ direction
+ elevation
+ quality
+ size
+ distance
+ intensity
+ ratio
+ color
+ distribution
+ modifiers
+ dynamics
+ atmosphere
+ continuity

8. The cut equation

CUT DECISION
= outgoing beat state
+ incoming beat state
+ motivation
+ boundary timing
+ visual relation
+ audio relation
+ continuity cost
+ audience effect

9. The infographic equation

INFOGRAPHIC
= question
+ information relationship
+ evidence
+ visual encoding
+ comparison frame
+ annotation
+ motion order
+ legibility window
+ source display

10. Production invariants

1. Meaning invariant — trimming must not reverse, exaggerate, or materially distort a speaker's meaning.
2. Provenance invariant — every source and generated asset retains origin and modification history.
3. Time invariant — source time and output time remain traceable.
4. Continuity invariant — intentional discontinuities are labeled as intentional decisions.
5. Purpose invariant — every cut, graphic, sound, or generated shot has a named function.
6. Legibility invariant — important information remains visible or audible long enough to be understood.
7. Evidence invariant — high-impact factual claims carry evidence state and verification status.
8. Accessibility invariant — required captions, transcripts, descriptions, contrast, and non-audio cues are part of production, not an afterthought.
9. Non-destructive invariant — source files remain immutable; decisions are stored separately.
10. Human-gate invariant — sensitive, factual, likeness, and publish decisions are reviewable and reversible.

11. Why the system is volumetric

A conventional preset chooses one answer. A volumetric system represents the entire feasible decision space.

Candidate space
= Content states
× Narrative states
× Camera states
× Lighting states
× Cut states
× Graphic states
× Audio states
× Delivery states
× LAKA change states

Blind Cartesian expansion is too large. The engine therefore uses:

Generate → Constrain → Score → Diversify → Review → Learn

Hard constraints eliminate invalid candidates. Soft constraints rank valid candidates. Diversity selection prevents every high-scoring candidate from looking the same.

12. Universal decision test

Before any operation, answer:

WHAT object changes?
WHY must it change?
UNDER WHAT conditions?
TOWARD WHAT audience state?
BY HOW MUCH?
FOR HOW LONG?
WHAT ELSE will it affect?
CAN it be reversed?
HOW will success or failure be detected?