SYMBIONTES™

CORPUS / 03

Analogies on a short leash.

The programme thinks with images from materials physics, cell biology, thermodynamics and cybernetics. It also obliges itself to say, in every use, whether the image is an analogy, a candidate model, a hypothesis or a rule. This page gathers the theories in play and what each one has not yet proved.

Version 0.6.0 · updated 2026-10-04 · Reading and measurement cut: 4 October 2026, Bahia time. · Version history

T1

How the programme uses analogies

Do not confuse metaphor with technical capability.

Technical Constitution v0.1, principle 8 (22 Jul 2026)

RECORDED FACTThe bibliographic audit of 27 September 2026 applied a three-layer reading to cognitive metabolism: first the biological or physical analogy, with the primary source of the field of origin; then a candidate informational model, formulated in terms of the system's own items, flows and states; finally a testable metric, with what it measures and what it does not. Extending that reading to every borrowed term is a hypothesis of the programme, recorded in a paper halted at the novelty gate.

“Genetic code” and “living membrane” are analogies, not demonstrated mechanisms.

Reconciliation of the elastoplastic cycle (27 Sep 2026)
TermLayer 1 · analogyLayer 2 · candidate modelLayer 3 · metricStatus
Elastoplasticitymaterials mechanicsload σ, strain ε, threshold Y, hysteresisfive proposed measures, none measuredhypothesis
Membranecell biologyselective, traceable, revocable, proportional boundary— (rule, not metric)rule in force
Cognitive metabolismbiochemistryitems that enter, wait, are transformed and leave with a destinationLittle's law L = λW (candidate)candidate model
Homeostasisphysiologydetection of saturation, lack of synthesis, backlogno objective indicatorshypothesis
Coupling (ATP synthase)bioenergeticsresource + input with provenance + bounded work + verified outputrework and losses, not measuredcandidate not adopted
Organs of the loopphysiologypacemaker, senses, excretion, memory—declared analogy
Entropy and equilibriumthermodynamicsgrow and disorganise, organise to growno definition or unitbrand language
Fractalgeometrythe same question at macro, meso and micro scales—declared analogy
Symbiosis, lichenevolutionary biologyhost, symbiont and relationship as three thingssovereignty metrics (exit)principle + analogy

T2

Elastoplasticity

The programme's oldest and most contested idea. It was born on 18 July 2026 as a model of a boundary, became the brand's anchor concept in August, was downgraded to hypothesis in the literature review, gained a synthetic mathematical model dated 18 July and, in the reconciliation of 27 September, a negative result.

Origin: a boundary that stretches

RECORDED FACTOn 18 July 2026 the founder asked for “elasticity” in the lock separating private from public. The assistant formalised an Elastic Exposure Membrane, with levels E0–E5, one level up per cycle and automatic retreat. At a single “revise”, the model became an Elastoplastic Transition Membrane, with a five-component vector and a saturation rule: momentary saturation means observe; recurring means reduce load; with error or loss of traceability, freeze. The rule survived as the direct antecedent of the current norm of freezing automations after an incident.

Anchor: implode and explode

To show that the host […] is imploding and exploding — entropy and equilibrium all the time, growing and encompassing. True elastoplasticity.

Founder, 2 Aug 2026

RECORDED FACTThe technical reply of the same day recognised the sentence as “the soul of the brand” and demanded observable definitions: elasticity as the time to return to equilibrium after a shock; plasticity as the structural change retained after the shock. The canonical thesis kept the verdict: “entropy, implosion and explosion remain brand language until they have a definition and a unit”.

Hypothesis P10: candidate variables

VariableProposed operational definition (23 Aug 2026)
Contextual load σvolume, heterogeneity, conflict, duration and number of handoffs
Strain εinstruction loss, error, drift, leakage or rework
Elastic regionrecovery after load removal or reset
Threshold Ypoint at which residual damage remains
Plastic straincontamination that persists after reset
Hysteresisdependence on the path and order of exposures
Schematic load–strain curveHorizontal axis strain, vertical axis load. Elastic line up to threshold Y, plastic curve afterwards, unloading parallel to the elastic line leaving retained strain, and a hysteresis loop.strain ε (instruction loss, error, drift, rework)load σ (volume, heterogeneity, conflict, duration, handoffs)Ythreshold Y: beyond it, retained change (damage in P10; learning in cycle 010)elastic region · full recoveryplastic region · retained changeretained plastic strainunloading · recovery (return time)reload · hysteresis: path-dependentfive proposed measures (not measured)absorption · degradation during the shockrecovery · time to returnretention · useful change keptreversibility · cost to undotransferred learning · next shock
Figure 1. HYPOTHESISLoad–strain curve as the programme uses it: elastic region (full recovery), threshold Y, retained plastic strain and hysteresis loop. The five measures proposed by the thesis are annotated on the curve. No value is measured; the curve is schematic.

Five measures, no measurement

  1. Absorption: service degradation during the shock, adjusted for volume and demand.
  2. Recovery: time to return to the operating band.
  3. Retention: useful change kept without degrading safeguards.
  4. Reversibility: cost and time to undo a bad change.
  5. Transferred learning: improvement at the next comparable shock or host.

RECORDED FACTThe five measures were proposed on 2 August 2026 and form part of the text promoted to canonical on 6 September, with the instruction to record observable components in the ledger first and compose an index only later. By the cut of this corpus, no value had been measured. The thesis adds a caution about the one number that tends to be quoted: zero collisions after the lease regime only becomes evidence with the denominator of comparable concurrent opportunities.

Two senses of plasticity

RECORDED FACTOn 27 September 2026 a reconciliation noted that the programme used “plasticity” in two senses: in the elastoplastic cycle of July, plasticity includes persistent transformation and learning, possibly favourable; in hypothesis P10, plastic strain is contamination that persists after reset, hence damage. The difference is semantic and sits in the sources; the reconciliation proposed keeping both (hypothesis; no normative decision) and distinguishing three things: reversible adaptation, persistent change and change in the capacity to adapt (metaplasticity). The transversal theory remains unvalidated.

DECLARED ANALOGYThere is also a third, architectural use: elasticity means changing views, granularity and batches while preserving identity; plasticity means changing structure and contracts with migration and history. The document proposing it says, in the same sentence, that no physical measurement or scientific validation is implied.

Status

T3

Entropy and equilibrium

We grow and disorganise, then organise to grow and so disorganise again. It is a cycle.

Founder, authorisation of the literature phase (2 Aug 2026)

RECORDED FACTThe sentence is the intuition that precedes elastoplasticity and the Threshold Doctrine. On 28 September 2026 the founder restated it as a problem: the project was born doing, testing and failing; now it creates laboratories before the canonical; it matures, but stiffens; if everything is fractal (macro, meso, micro), how to prevent complexity from becoming worse than entropy?

RECORDED FACTThe recorded answer was structural, not thermodynamic: a barbell-shaped threshold, with free execution below and laboratory above, and three locks against fossilisation, one of them called “budgeted entropy”: every laboratory has an expiry, every gate adjusts itself, and there is an explicit budget of tolerated disorder. None of these locks has a calibrated parameter.

T4

Cognitive metabolism and homeostasis

RECORDED FACTThe term appears on 16 July 2026 as an inheritance between two founding cycles and was audited on 27 September. Definition in use: cognitive metabolism as a sequence of capture, relation, synthesis, consolidation, publication, archive and redundancy reduction. The model cycle is expansion → deepening → synthesis → return → consolidation, accompanied by maturity states and locks.

Cognitive metabolism as a queueSeven stages in sequence, from capture to redundancy reduction, with items entering and leaving; Little's law noted below, with what it does not measure.λcapturerelationsynthesisconsolidationpublicationarchiveredundancy reductionL · mean stock of items in processW · mean time through the systemL = λ · Wcandidate metric · does not measure truth, synthesis, quality or value · document flows are not physico-chemical reactionsmodel cycle: expansion → deepening → synthesis → return → consolidation · biological analogy; candidate informational model
Figure 2. CANDIDATEThe metabolic cycle as an informational model: items enter, wait, are transformed, consolidated and leave with an explicit destination. Little's law relates mean stock, arrival rate and time through the system. What it does not measure is written on the figure.
L = λ · W
L
mean stock of knowledge items in process
λ
effective arrival and departure rate of items
W
mean time an item takes to pass through

CANDIDATELittle's law is the candidate metric. The audit records what it does not do: it does not measure truth, synthesis, quality or value. And it records the limit of the analogy: document flows are not physico-chemical reactions; no source consulted shows the symbiont to be a living organism or an autopoietic system. The safe formulation is “cognitive metabolism: biological analogy; candidate informational model”.

Homeostasis

HYPOTHESISFirst occurrence located in the fifth founding notebook: the need to recognise excessive expansion, lack of synthesis, backlog and the moment to close. Declared state: a mature hypothesis, with evidence only of human detection of saturation and of transition failures. Without objective indicators, it is assisted diagnosis, not regulation.

Maturity states and verb locks

RECORDED FACTThe same notebook proposed a ladder of states for ideas: insight, hypothesis, experimental principle, operational decision, constitutional rule, doctrine; and a textual lock, “reading does not authorise altering”, which separates acts by verb. The ladder is used; the passage criteria between rungs are still absent. None of the notebook's proposals was promoted: Constitution v0.4 remains the only one in force.

T5

Operational physiology: organs

DECLARED ANALOGYIn August 2026 the programme drew the host company as an organism of organs, each with sensor, controller, effector, feedback, deficit and dependency. The example organ, named the mitochondrion, would have the vital function of converting cash, time, attention and professional capacity into sustainable work per service order. The record itself states the condition: a calculable design exists; the metabolism linked to real operation is missing; the formulas are not fed by events.

RECORDED FACTThe documentary lineage goes from a “periodic table” of the operation's elements, bonds and reactions to a passage from chemistry to systems biology and, finally, to a register of operational organs. It is a sequence of increasingly systemic analogies; the entry read is marked as designed, not fed.

INFERENCEThe September 2026 Atlas translated the office into a body: a brain that decides, hands that do, nerves that carry the order, automatic posts as a heart, AI models as mitochondria, bases as skeleton and memory. The current edition qualifies the image: the October direction treats the two cores as peers, neither being the other's brain.

T6

ATP-synthase-inspired coupling

The founder's question on 1 October 2026: can ATP synthase help with anything in our architecture? A mission was opened to answer, with the analogy kept under control.

RECORDED FACTThe biological mechanism has primary sources: the structure of the enzyme's catalytic portion (1994), the direct observation of its rotation (1997) and proton-powered rotation in the whole membrane-bound enzyme (2004). They support the physical coupling between electrochemical gradient, rotation and catalysis. None of this is carried over literally.

Verifiable couplingInput with provenance passes an admission gate, drives a rotor of bounded work fed by available resource, and yields verified output; losses listed below; on the right, the list of what the analogy does not carry over.coupled inputevidence with provenance and contractadmission gateprovenance? mandate? no order from contentworkboundedavailable resource (capacity, time; never authority)verified outputreceipt + readback at destinationlossesduplication · return without key · rework (not measured)does not carry over— data is not a proton or energy— token budget is not a gradient— rotor is not a scheduler— three conformations ≠ three queue states— ATP does not measure profit— biochemical reversibility ≠ undoing a human decisiondesign hypothesis; candidate not adopted; synthetic trial 3/3 = internal consistency, not efficacy
Figure 3. CANDIDATEVerifiable coupling as the programme translated it: available execution resource, input with provenance and contract, bounded work and verified result; losses are duplication, return without key and rework. The list of what the analogy does not authorise sits alongside.
In the enzymeIn the programme (translation)What does not carry over
Proton gradientresource = capacity and time available to process a step; never authoritydata or evidence is not a proton or energy; token budget is not a gradient
Rotorwork bounded by contractthe rotor is not a task scheduler
Three catalytic conformations—do not authorise imposing three states on the queue
ATPuseful work = bounded output, delivered and verifiedATP does not measure profit
Reaction reversibility—biochemical reversibility does not legitimise undoing human decisions

The biological analogy is a design hypothesis, not a software mechanism nor a source of authority.

Coupling architecture mission (3 Oct 2026)

T7

Agentic loop and the four organs

The true loop: S3/S4 pull H1/H2 → build H3/H4 → feed S1/S2 → improve S3/S4 — a circuit above every other process in the system.

Founder, 24 Sep 2026 (excerpt)
The double loopFour host vaults on the left and four symbiont vaults on the right, joined by four numbered arrows; above, reality and a dashed result signal, never specified.REALITY · who answers · the outcome out thereexternal result signal: never specifiedHOSTSYMBIONTH1 · context and casesH2 · playbooksH3 · capabilitiesH4 · agentic companyS1 · hosting casesS2 · hosting playbookS3 · agentsS4 · coordination12341 · S pulls context and method from H · 2 · builds H's capabilities · 3 · feeds S's memory · 4 · improves S's capabilitiessolid: exists to some degree · dashed: does not yet exist
Figure 4. RECORDED FACTThe double loop: the symbiont's capabilities pull context and method from the host, build the host's capabilities, feed the symbiont's memory and improve their own capabilities. Above the loop, the reality it does not rewrite. The external result signal remains dashed: it was never specified.

INFERENCEThe diagnosis of 24 September 2026 identified four organs without which the loop does not close: a pacemaker (scheduled execution; without it there is a sequence of sessions, not a loop); senses (an external result signal returning to the loop; without it the cycle repeats but does not learn from the real outcome); excretion (closure and archiving rule); and the symbiont's memory (hosting cases and hosting playbook).

The four organs of the loopA central ring represents the loop between symbiont and host; four organs around it: pacemaker, senses, excretion and memory, each with a state. Senses is dashed: not covered.LOOP S ↔ Hcontext → method → capability → coordinationPACEMAKERscheduled executiondelivered 27 Sep · daily routine; 9 hourly watches fired 4 OctSENSESexternal result signalNOT COVERED · never specifiedEXCRETIONclosure and archivingabsorbed by the closure rulesMEMORYhosting cases and playbookdelivered 26 Sep · two S-side baseswithout the pacemaker there is a sequence of sessions, not a loop; without senses the cycle repeats but does not learn from the real outcome
Figure 5. RECORDED FACTState of the four organs on 27 Sep 2026: memory delivered (two S-side bases), excretion absorbed by the closure rules, pacemaker delivered as a daily routine, senses not covered. On 4 October the pacemaker had been complemented by nine hourly watches, one per vault, with first run observed.

CANDIDATEThe pacemaker architecture was revised on 3 October 2026: the AIs already wake on their own; what is missing is a clock that depends on nobody's session, that records every attempt before acting, that does not repeat work and that proves on its own what it did. A synthetic trial was concluded on 4 October; the report keeps two caveats from the architecture document (v0.2): execution success does not prove effect, and cloud independence was not demonstrated. Nothing was deployed.

Status is a label; capability is a membrane.

Pacemaker and ten-vault architecture, v0.2 (3 Oct 2026)

T8

Fractality

HUMAN DECISION“Everything is fractal” is a recorded decision of the founder on 26 September 2026. Constitution v0.4 already fixed three analytical scales (macro, meso, micro). The September update added the relational fractal: unit, sector, vault, service, case, document and mission are different axes; matrices are views over relations, not a mandatory physical tree.

slicing criterion (27 Sep 2026)
create_logical_slice(x)      := changes_the_question(x) ∨ changes_the_owner(x) ∨ changes_the_acceptance_criterion(x)
create_physical_container(x) := access(x) ∨ volume(x) ∨ lifecycle(x) ∨ operation(x) justify it

void ∈ { missing_inventory, confirmed_non_existence, non_applicability }   // three kinds, never confused
identity := immutable ; classification := mutable
Change of scaleAt scale zero, SH coordinates S and H inside Q zero. At scale one, the whole Q zero becomes H one and SH zero becomes S one.SCALE 0SH₀Q₀ = { S₀, H₀, bridges, membrane }change of scaleSCALE 1SH₁ · only with a new compositionS₁ = SH₀H₁ = Q₀the inner drawing is preserved · the rule can repeat but needs a limit (in progress)
Figure 6. RECORDED FACTThe scale recursion drawn on 2 October 2026: at the next scale, the whole square occupies H and the previous coordination occupies S. Two scales are the graphic limit; the operational limit is in progress.
Three bridgesAn S region with an internal bridge, an H region with an internal bridge and, between them, the SH bridge with a membrane.S · symbiont sideinternal bridge Squeue, current state, S-side vaultsH · host sideinternal bridge HH-side team and AIs, entry portal, H-side vaultsSH bridgeshared missions · common portalmembraneaccess: Central → S, H and SH · Operational → H and SH · partial implementation; effective separation not declared without proof of permissions
Figure 7. RECORDED FACTThree bridges, a direction of 1 October 2026: the internal bridge of the S side, the internal bridge of the H side and the SH bridge between them, the only one reachable from both sides. Partial implementation; effective separation is not declared without proof of permissions.

DECLARED ANALOGYA diagram from August 2026, with no record promoting it, states a fractal “fundamental law” with four flows (decision descends, evidence ascends, context crosses, resources descend) and seven principles: isolation by default, access by need, evidence before conclusion, ethics above all, escalation on doubt, regression allowed and “leaps only with MOR”, an abbreviation with more than one meaning in the archive, read here as human review. The programme treats it as a visual model: useful for governance, with no proof of a universal law.

T9

Symbiosis as a relationship, not an organism

RECORDED FACTThe central thesis of the September Atlas: a symbiont is not a symbiosis. Host, symbiont and relationship are three words that are not synonyms, and the value lies in the design of the relationship, not in the strength of either pole. The lichen is the didactic image: two species, one composite organism, neither explaining it alone.

INFERENCEThe Atlas's philosophical reading applied four references to the system: symbiogenesis (Margulis), organic solidarity (Durkheim), the iron cage of bureaucracy (Weber) and the responsibility of whoever has a body (Jonas). The e-book marks them as the Central's inference, a didactic translation, not a replacement for the architecture documents.

T10

Cybernetics, information and the translation canon

HYPOTHESISOn 17 July 2026 the programme fixed a three-level translation canon: original theory → structural principle → testable rule, citing Einstein, Newton, Maxwell, Tesla, Shannon, Wiener, Ashby, Prigogine, Darwin, Turing and Simon. The document is classified as a hypothesis; “canon” in the title does not give it canonical status.

RECORDED FACTTwo demonstrations were attempted on 27 September 2026: Shannon, with bounded bibliographic support; Einstein, inconclusive for lack of access to the text. Tesla is cited for electrical systems and resonance; a numerical pilot of July (the sequence 3, 33, 333) was recorded as a coincidence not promoted.

RECORDED FACTThe cybernetic vocabulary is in use: continuity radar, sensor, kill switch, control plane, homeostasis. The corresponding computational mechanisms (lease, idempotency, heartbeat, hashes, logical clock, branch identifiers) are implemented and tested; they are the least metaphorical part of the system.