The Living Video
The Living Video
See also: The Living Book, The Living Podcast, The Living Papers, The Living System, Harmonism.
Most videos are timestamped — produced once, frozen at the moment of the cut. A philosophy that claims to describe a living reality deserves a living visual transmission.
The Living Video is the video version of the harmonism.io corpus, paired from canonical-article audio and a curated visual layer that composes from a hand-shaped asset library. Each video corresponds to a canonical article. When the article deepens in the vault, the video deepens with it — the audio regenerates from the updated text via the TTS pipeline, the visual layer recomposes against the updated audio, and the version a practitioner watches today is the current doctrinal version. No talking head, no founder face, no frozen-at-recording-date drift.
This is the third reception layer of the corpus. Text on the page, audio in the ear, video in the eye — three faces of one articulation, all sourced from the same canonical text, all regenerating together.
What it is
A library of videos, one per canonical article, distributed via YouTube and harmonism.io’s own browsable archive. The audio track is the same TTS narration that powers The Living Podcast; the visual layer composes from a curated asset palette to give the eye something coherent to follow while the ear receives the prose.
The visual palette is hand-designed and additive over time. It includes the full Wheel of Harmony SVG and its eight sub-wheels (already in the codebase, used in the website’s interactive layer); the Architecture of Harmony hendecagon (already in the codebase); the page essences that carry each article’s compressed seeing in poetic prose (already curated for ~250 articles); the classification badges that mark each article’s content layer, doctrinal status, breadth, depth, and craft; the glossary terms that surface as text overlays as the narration reaches them; country and figure imagery for Country articles and People articles; the visual identity of the Harmonist project (the heptagon logo, the plum and indigo palette, the typography system). The orchestration layer pairs visuals to audio timecodes per article so the visual unfolds with the prose rather than running independently of it.
What makes it living
The same Living-X discipline that governs The Living Book and The Living Podcast governs The Living Video. The canonical article is the source of truth. When the article changes, the audio regenerates, the visual recomposes against the new audio, the published video updates. A practitioner who watches Harmonism six months from now sees the current articulation — including any doctrinal refinements that have landed since the last viewing.
This refuses what most video pipelines do. A talking-head video can’t update — the recording freezes the content at the moment the camera stopped, and any subsequent doctrinal refinement requires re-recording the entire piece. Re-recording at corpus scale is operationally impossible. The Living Video pipeline solves this structurally: the canonical article is the only thing edited; everything downstream regenerates.
Why no human face
Decision #637 names the founder only in the About the Founder article. Talking-head canonical-article video would put a visible founder identity on system-voice content and contradict the doctrine. The Living Video keeps the system speaking through the prose without a human face attached. The visual layer is the Harmonist visual register — the Wheel, the Architecture, the page essences, the diagrams, the typography — composed in service of the audio narration.
The exception is the The Harmonia Podcast, the guest interview show, where Tahir-in-conversation-with-substantive-figures is the right form because the encounter itself is the content. The podcast is its own production track, structurally separate from the Living Video pipeline.
Where it lives
The browsable page on harmonism.io carries the full library — video cards with thumbnails (a still frame from the visual layer, typically a Wheel or Architecture composition that signals the article’s structural location), titles, brief descriptions, embed for direct play, and a link to the article page where the full text, audio, and video live together. The same videos are mirrored to YouTube under the institutional channel for discovery via search and recommendation.
Phase
Architecturally committed and build-deferred. The visual asset library exists in the codebase. The TTS audio pipeline is operational. The orchestration layer that pairs visuals to audio timecodes per article is the missing piece — that is the Phase B build. Phase A: scaffold (this article, the pipeline doc, the asset library inventory, the architectural commitment recorded in the Decision Log). Phase B: build the orchestration layer, render the keystone articles as video, deploy the browsable page, mirror to YouTube, integrate with the practitioner subscription invitation.
When the orchestration layer ships, the rendering is automatic — every canonical article that has TTS audio gets a corresponding Living Video without separate production work. The corpus releases at three reception layers from one source.