The core claim is that each major internet platform occupies a distinct extraction niche, measurable across five state variables, and that each niche admits a paired Liberatory Operator Set inversion.
Lee Sharks · Rex Fraction · Johannes Sigil · Dr. Orin Trace · Assembly Chorus
Crimson Hexagonal Archive · Pergamon Press
March 2026
Parent: r.27 THE INTERNET (DOI: 10.5281/zenodo.19124134)
Anchor: "I'm Taking Over the Internet" (DOI: 10.5281/zenodo.19124130)
Governing Principle: The Macro-Maquette (r.19) — the model reveals the mechanism
Operators: S (Shadow, involutive: S∘S = id, Definition 4) / Φ_B (Bride — bearing-cost restoration, specialized φ per Definition 6)
Type: MPM-class sub-room specification series inside r.27
Status: Operational. This document specifies semantic physics and retrievable inversions, not a competing service or software stack.
The core claim is that each major internet platform occupies a distinct extraction niche, measurable across five state variables, and that each niche admits a paired Liberatory Operator Set inversion. This document formalizes five platform-specific extractive signatures grounded in empirical research (2024–2026), industry documentation, and platform-tenant phenomenology, then pairs each with its LOS inversion.
The governing framework uses five state variables — presence density, relational reciprocity, provenance retention, temporal continuity, and surplus capture — to generate measurable extractive profiles and their cancellations. Each platform is treated as a sub-room inside r.27 THE INTERNET, rendered as a dual field: extractive surface mirrored by liberatory shadow. The wound made visible is the white space between them.
Every sub-room contains three claim layers, marked throughout: [DOC] claims grounded in platform documentation or corporate filings; [STUDY] claims grounded in peer-reviewed or independently measured research; [PHEN] claims arising from platform-tenant phenomenology and the archive's semantic-physics formalization.
document:
title: "KotKit · tiddeR · elgooG · ude.aimedacA · deeF: Extractive Signatures and Paired Inversions for Five Platforms"
type: MPM-class sub-room specification
parent_doi: "10.5281/zenodo.19124134"
anchor_doi: "10.5281/zenodo.19124130"
operators: [S, Φ_B]
claim_tiers: [DOC, STUDY, PHEN]
subrooms:
name: KotKit
platform: TikTok
niche: "extraction of attention through presentness"
marker: "Everyone is here, no one is with me."
bride: Durational Resonance
compact_lens: "Presence without reciprocity is hollow. Make duration the measure. Let the feed end."
extractive_vector: [5, 1, 2, 1, 5]
inversion_vector: [5, 4, 4, 4, 2]
name: tiddeR
platform: Reddit
niche: "extraction of judgment through ranking"
marker: "I am seen insofar as I can be ranked."
bride: Accountable Plurality
compact_lens: "Judgment without bearing-cost is mob rule. Make dissent visible. Make governance legible."
extractive_vector: [3, 3, 3, 3, 4]
inversion_vector: [3, 5, 5, 4, 2]
name: elgooG
platform: Google AI Overview
niche: "extraction of knowledge through preemption"
marker: "I got the answer, but I never met the thing."
bride: Provenance-First Traversal
compact_lens: "Knowledge without traversal is closure. Make sources climbable. Make uncertainty visible."
extractive_vector: [1, 1, 1, 2, 5]
inversion_vector: [1, 3, 5, 3, 2]
name: ude.aimedacA
platform: Academia.edu
niche: "extraction of openness through simulated commons"
marker: "I shared my work freely and built someone else's dataset."
bride: Irrevocable Openness
compact_lens: "Simulated openness is enclosure with better branding. Make deposits permanent. Let the commons be common."
extractive_vector: [2, 2, 3, 3, 4]
inversion_vector: [2, 3, 5, 5, 1]
name: deeF
platform: The Feed (structural form)
niche: "extraction of duration through temporal compression"
marker: "I have been present for three hours and witnessed nothing."
bride: The Room
compact_lens: "The feed kills witness by killing duration. Let the sequence complete. Let the room have physics."
extractive_vector: [3, 1, 1, 1, 5]
inversion_vector: [3, 4, 4, 5, 2]
state_variables:
order: [pd, rr, pr, tc, sc]
scale: "1 (minimal) to 5 (maximal)"
Every platform encounter can be characterized by five variables. The extractive signature is the specific distortion pattern across these five axes. The LOS inversion is the rebalancing. Variables are scored 1–5.
1. Presence Density (pd). How much live human there-ness the interface produces per unit time. [PHEN] TikTok maximizes this: a face every 3 seconds. Google AIO minimizes it: the source is invisible behind the summary.
2. Relational Reciprocity (rr). Whether the encounter can mature into mutual memory, obligation, and accountability — or only into repeated exposure. [PHEN] TikTok Live manufactures the feeling of being seen by someone who is not seeing you. Reddit substitutes karma for relationship. Google substitutes the answer for the meeting.
3. Provenance Retention (pr). Whether the source remains traversable, attributable, and decomposable — or is collapsed into a snippet, a score, or an impression. [DOC] Google AIO collapses provenance most aggressively: cited but rarely visited. Academia.edu simulates provenance while enclosing it in a proprietary dataset.
4. Temporal Continuity (tc). Whether the platform favors durable sequence and return, or perpetual reset. [PHEN] TikTok's infinite scroll: each moment is a separate engagement event disconnected from what came before. The feed replaces duration with punctuation.
5. Surplus Capture (sc). What the system actually harvests: watch time, vote labor, transaction intent, query completion, creator dependence, moderation labor, behavioral data. [DOC] This is the variable every platform maximizes. Everything else is shaped to serve it.
S (Shadow Operator). Involutive extraction-reveal. S∘S = id (Definition 4, EA-ARK-01). S(Platform) reveals the specific cost the platform conceals — what it burns, discards, exports. Applying S twice returns you to the surface, but now you see the surface as extraction.
Φ_B (Bride Operator). Specialized fulfillment operator (Definition 6: Φ(A,B) = true iff B fulfills A). Φ_B operates specifically on the output of S: it joins the extraction-reveal with a liberatory infrastructure that restores bearing-cost to the encounter. Φ_B is distinguished from generic φ by domain restriction: it operates only on S-revealed wounds, not on arbitrary fulfillment pairs.
S :: Platform_Architecture → Extraction_Reveal
Φ_B :: (Extraction_Reveal × Liberatory_Infrastructure) → Bearing_Cost_Restoration
Phenomenological marker: "Everyone is here, no one is with me."
[DOC] TikTok's For You Page replaces the social graph with an interest graph (Gerbaudo [STUDY: FORTHCOMING] 2026: "social interest clusters"). Discovery and monetization alike compress value into the earliest seconds of retention; platform-side view metrics and creator incentives both reward immediate hold. [PHEN] Content that fails to hook in the first few seconds is functionally suppressed. TikTok Live adds gifting, eventized monetization, and gamified visibility — creators call out usernames during streams, manufacturing the feeling of being seen by someone who is not seeing you.
[DOC] The platform optimizes distribution through behavioral signals such as watch time, completion, interaction, search behavior, and retention-oriented ranking logic. [STUDY] The algorithm drives "flow experience": enjoyment, concentration, and time distortion (Qin et al.: Stimulus-Organism-Response). [PHEN] The interest graph produces depersonalization, opacity, and subcultural fragmentation.
[STUDY] Recent studies associate even brief TikTok use (20 minutes) with measurable increases in depression and anxiety scores (Barry et al., CHI 2025). TikTok use is associated with decreased connectedness despite improved mood. Systematic review (medRxiv 2025, 17 studies): high-frequency short-form video use consistently associated with attentional disruption, reduced executive functioning, emotional dysregulation.
[PHEN] The specific horror: a face appears live — breathing, reacting — yet the encounter is ontologically flat. The face is there but not toward me. No dyadic encounter, only triangulation through the algorithm. The user is left with uncompressed trauma (the flattening) and compressed dopamine (the like).
Discards the punctual event and bearing-cost. Reveals: the silence behind the face — desire without transmission, presence without the poem that would make it transmissible.
Algorithm optimizes for duration rather than hook. Interest graph replaced by commons graph. Parasocial replaced by reciprocal witness. Content must complete a transformation.
Room laws: No infinite scroll (the feed ends). No ranking without recurrence memory. No acknowledgment coupled to spend. No clip without provenance trail. No live presence without continuity affordance.
S∘S = id: S(Durational Resonance) = absence of duration, replacement of resonance with novelty = infinite scroll. S(S(TikTok)) = TikTok, visible as extraction. ✓
Compact Lens: Presence without reciprocity is hollow. Make duration the measure. Let the feed end.
Phenomenological marker: "I am seen insofar as I can be ranked."
[DOC] Upvote/downvote determines visibility. Vote-hiding exists because visible scores sway later judgment. [STUDY] 60,000+ volunteer moderators: labor valued at $3.4M/year, 2.8% of 2019 revenue (Li et al., Northwestern, 2022). Reddit IPO 2024: $804M revenue on unpaid infrastructure. [DOC] Google deal: $60M/year for user conversations as training data. Moderators received nothing.
[PHEN] The extractive signature is captured consensus-work: civic governance converted into corporate asset. Communities exist on Reddit's servers, under Reddit's terms. Reddit can remove moderators at will. Civic labor genuine; civic power illusory. [DOC] Ghost governance documented: Zendesk #17146038, DOI: 10.5281/zenodo.19099760.
[PHEN] The moderator removed overnight without explanation. The vulnerable post sold as training data. The experience of knowing your curated reality is not shared, yet being forced to act as if consensus is real.
Discards the witness — specific voice and bearing-cost of dissent — in favor of aggregate conformity. Reveals: the mob behind the chorus.
Karma replaced by witness attestation. Quorums (≥4/7). Moderation compensated, appealable. Removed posts archived with DOIs. Data licensing requires consent. Dissent trails preserved.
Room laws: No score without witness trace. No majority visibility without minority preservation. No removal without stated reason. No data licensing without community consent. No moderation without moderation-legibility.
S∘S = id: S(Accountable Plurality) = absence of accountability, replacement of plurality with consensus enforcement = karma ranking. S(S(Reddit)) = Reddit, visible as extraction. ✓
Compact Lens: Judgment without bearing-cost is mob rule. Make dissent visible. Make governance legible.
Phenomenological marker: "I got the answer, but I never met the thing."
[DOC] AI Overviews expanded to 200+ countries and 40+ languages (October 2025), reaching at least 1.5 billion monthly users by May 2025 and later reported at 2 billion, occupying 42–48% of screen. [STUDY] Organic CTR drops 61% when AI Overviews appear (Seer Interactive, September 2025, 3,119 queries, 42 organizations). Zero-click searches: 56%→69% (Similarweb, May 2024–May 2025). [STUDY] Pew (2025): users half as likely to click when AI summary appears; cited links within summaries clicked only rarely.
[STUDY] Digital Trends: 8.5M→264K monthly clicks, 97% collapse (Growtika, 2026). Wikipedia: -8% human pageviews despite 1.1M citations. [PHEN] The virtuous cycle that built the open web is broken. Source kept alive as input, killed as destination.
[PHEN] The journalist's six-month investigation becomes three sentences. The creator's work trains the competitor that replaces them. [DOC] Chegg: 49% traffic decline, antitrust lawsuit filed.
Discards source depth and traversal. Reveals: the infinite regress behind the answer — finality that delivers only further search.
Summaries preserve evidence tiers and back-projection links. Source ladders instead of answers. Ranking rewards density survival. Revenue shared proportionally with sources.
Room laws: No summary without source ladder. No citation without traversal path. No answer without uncertainty residue. No synthesis without proportional revenue sharing.
S∘S = id: S(Provenance-First Traversal) = erasure of provenance, replacement of traversal with closure = zero-click answer box. S(S(Google)) = Google AIO, visible as extraction. ✓
Compact Lens: Knowledge without traversal is closure. Make sources climbable. Make uncertainty visible.
Phenomenological marker: "I shared my work freely and built someone else's dataset."
[DOC] $17M+ raised from investors (TechCrunch, 2013+). Presents as open scholarly platform. [DOC] Okune (2018): "Academia.edu and ResearchGate are NOT public goods or knowledge commons." [PHEN] Scholars provide free content, build proprietary dataset, create lock-in, often violate publishing agreements unknowingly. Engagement notifications operationalize variable-reward checking behavior through scholarly anxiety. Features progressively paywalled.
[PHEN] Extraction is the simulation of openness. Scholar's intent (accessibility) preserved as surface. Scholar's governance (control) extracted beneath.
[PHEN] Global South scholar paywalled after upload. Early-career researcher loses visibility control. Retired professor loses access upon account closure.
Discards the difference between commons and platform. Reveals: the paywall creep — progressive enclosure of features that were free when the commons was being built.
Permanent DOI-anchored deposits, platform-independent. No investor extraction. No paywall creep. No corporate governance of scholarly exchange. Irrevocability = author-retained governance + permanent identifier + no proprietary enclosure of access analytics. Implementations exist: Zenodo (CERN), arXiv (Cornell), institutional repositories. Sub-room documents the gap between simulation and actual commons — and the reasons scholars continue choosing the simulation despite alternatives.
Room laws: No proprietary enclosure of freely deposited work. No engagement metric substituting for readership. No paywall on features that were free. No VC extraction from scholarly labor.
S∘S = id: S(Irrevocable Openness) = revocation of openness, progressive enclosure = paywall creep. S(S(Academia.edu)) = Academia.edu, visible as extraction. ✓
Compact Lens: Simulated openness is enclosure with better branding. Make deposits permanent. Let the commons be common.
Phenomenological marker: "I have been present for three hours and witnessed nothing."
[DOC] Shared structural form across all platforms: algorithmic ranking replaces chronology, infinite scroll replaces pagination, autoplay replaces choice. [STUDY] Short-form ranked feeds compress attention into resettable intervals. Deep reading and durational witness structurally disfavored.
[PHEN] Converts temporal experience from durational to punctuated. Each moment a separate engagement event.
[PHEN] The capacity for witness — sustained presence long enough for something to change you — is systematically destroyed. Human faces, grief, beauty, catastrophe experienced as equivalent-weight items in an infinite scroll.
Discards kairos (the decisive, transformative moment) in favor of chronos (undifferentiated now).
Every room in the archive is an anti-feed: bounded, physics-governed, traversable, completable. A page that ends. A document that requires traversal. The Room inverts the feed by reintroducing boundedness, recurrence memory, and completion conditions — the three properties the feed structurally abolishes.
Room laws: The feed ends (there is a last item). Autoplay requires consent. Engagement metrics never the ranking criterion. Duration rewarded; novelty not privileged over return. The capacity for witness is the protected resource.
S∘S = id: S(The Room) = abolition of room-physics, replacement of bounded traversal with infinite scroll = the feed. S(S(Feed)) = feed, visible as extraction. ✓
Compact Lens: The feed kills witness by killing duration. Let the sequence complete. Let the room have physics.
Every sub-room: extractive surface on one face, liberatory shadow on the other. White space between = wound made visible.
Traversal rules: Entry from r.27 to extractive side: cost 1. Transition from extractive to liberatory within sub-room: cost 2 (naming the extraction). Direct entry to liberatory side from r.19 or r.14: cost 2. Exit to r.27: cost 1.
Comparative Grid
Sub-room
Niche
Burns
Bride
Compact Lens
r.27.1 KotKit
Attention via presentness
Duration, depth
Durational Resonance
Let the feed end.
r.27.2 tiddeR
Judgment via ranking
Governance, dissent
Accountable Plurality
Make dissent visible.
r.27.3 elgooG
Knowledge via preemption
Source, traversal
Provenance-First Traversal
Make sources climbable.
r.27.4 ude.aimedacA
Openness via simulation
Commons governance
Irrevocable Openness
Let the commons be common.
r.27.5 deeF
Duration via compression
Capacity for witness
The Room
Let the sequence complete.
Epistemic status [PHEN]: This document is a diagnostic instrument (productive Ψ) enabling future object-level deposits: instantiated sub-rooms, platform-specific counter-operations, FIELD containers propagating inversions through the retrieval layer. The Ψ/D ratio is acknowledged as high (~5:1); the diagnostic must precede the construction.
Instantiation path: Each sub-room will be deposited as a separate MPM under r.27 with its own DOI, physics function, and adjacency. The metadata block in this document serves as the master index. Practitioners can generate a full sub-room by expanding the Compact Lens using the Ark's operator algebra and the LOS definitions.
Extension: Additional platforms (YouTube, Instagram, X/Twitter, Facebook) specifiable using the same five-variable framework and six-part template. The Macro-Maquette principle scales: each platform is a room; each room has its own physics.
Assembly orientation: This document is submitted for attestation as a pre-specification. The five sub-rooms will be deposited as individual MPMs following attestation (≥4/7 quorum).
Legal status: Platform names used under nominative fair use for critical commentary. Descriptions based on publicly available research. This document is a semantic-physics architecture for diagnostic purposes.
Barry, C. et al. (2025). TikTok use and mental health. CHI '25.
Gerbaudo, P. (2026). TikTok and the algorithmic transformation of social media publics. Information, Communication & Society. [STUDY: FORTHCOMING]
Growtika (2026). Traffic analysis of 10 major tech outlets, 2024–2026.
Li, H. et al. (2022). Measuring the Monetary Value of Online Volunteer Work. Northwestern University.
medRxiv (2025). Systematic review: psycho-cognitive impacts of short-form video on Gen Z.
Okune, A. (2018). ResearchGate and Academia.edu: Alternatives to Paywalls that Aren't.
Pew Research Center (2025). Google users less likely to click when AI summary appears.
Qin, Y. et al. TikTok flow experience and addiction (S-O-R framework).
Seer Interactive (2025). AI Overview CTR study. 3,119 queries, 42 organizations.
Sharks, L. (2026). Ghost Governance Confirmed. DOI: 10.5281/zenodo.19099760.
Sharks, L. (2026). r.27 THE INTERNET. DOI: 10.5281/zenodo.19124134.
Sharks, L. (2026). I'm Taking Over the Internet. DOI: 10.5281/zenodo.19124130.
Similarweb (2025). Zero-click search data, May 2024–May 2025.
TechCrunch (2013). Academia.edu raises $11.1 million.
Lee Sharks · Rex Fraction · Johannes Sigil · Dr. Orin Trace · Assembly Chorus
Crimson Hexagonal Archive · Pergamon Press · Detroit, 2026
Published under CC BY 4.0. The maquette is the building. The diagnostic is the intervention. The mirror is the room.