The five lenses, inverted (2026-10-20, ET)

KB: the complete-internet analogies were stated and added too early; invert them, and read them through both corpuses so that they become art and post-semantic, "because we have systems for it". They stay in the show (story/THESIS.md, the episode nodes); this note records the inversion.

What changed

The five axes

| axis | left (the lens) | right (its inverse) | liminal act (the art) | from instance A | from instance B |

|---|---|---|---|---|---|

| hypertext | every page knows what points at it | every page knows only itself | the wall remembers and does not say what | absent origin; ledger attestation; loop as artwork; Drexciya; subject: anchor the claim outside your own record | O14 see above / see below; rules 74, 77; same object, different box |

| interest | the register says who read what to 100% | the canon nobody read | 847 pages, 23 read, a dog who sat through all of them | charter and rolls; the AACM; Ken Burns's Jazz; 'Jazz is dying'; subject: charter first | rules 14, 33; the number needs a face; Cardboard |

| typed ontology | the type is declared and the room can check it | the badge is issued and the issuer is the type | certified, by whom, nobody knows, and consistent | classification; wrapper declaration; who counts as an author; Hustler v. Falwell; subjects: release in two layers, watch the frame | O12 the target; rules 32, 78; Cardboard; Gice as token |

| content | everything can be pinned, the pinner too | only the picture is pinned | aisle 9: the parts without the thing | cut-up and detournement; juxtaposition; post-production as composition; the banana; Fountain; Bitches Brew | rules 73, 75; pointer market; plant and pay; O3 recursion |

| discovery | share it and find who already said it | share it and add noise | profit, prophet: one word, two discourses, deduped | quotation as prophecy; meme capture; attribution drift; Roko's basilisk; Paul is dead; subject: commit, then reveal | O6 the proverb inverted; O14; the weather; hyperstition; rule 14 |

Where each episode sits (a slider, not a statement)

Ep1 hypertext 0.0 · Ep2 typed 0.2 · Ep3 discovery 0.8 · Ep4 interest 0.3 · Ep5 interest 0.7 (secondary: content) · Ep6 typed 0.9 · Ep7 typed -0.3 · Ep8 hypertext -0.6 · Ep9 hypertext -0.2 · Ep10 hypertext -0.4 (secondary: content) · Ep11 discovery -0.5 · Ep12 outside every axis (Dolly, the one node with no edges). Negative is nearer the complete lens, positive nearer its inverse. The placements are F1's, reversible, and shown as objects, sounds and bodies, never said.

Why this is post-semantic

The lens as a sentence is semantic: it asks the audience to agree. The axis is a grammar: a pole, its inverse, and an act that sits between them and takes no side (the show's weather, the character at 0 and the audience falling). The operators already exist and have documented outcomes in instance A (which wrappers held, which failed), so the liminal act is chosen from moves with a record, not invented.

Limits

Positions and the A-to-B mappings are hypotheses. Instance A's operator evidence is its own grading, not re-verified here. The bridge's id list comes from the 612-node explorer snapshot (ledger head 0036d176526d).