The Three Axes They Kept Off the Chart

Our own political-taxonomy frame for the Evil Robots series and evilrobots.lol — NOT an adopted external model (Pew/8values/etc. import their own loaded labels).

2026-06-07 14 min read Research file
Contents

Our own political-taxonomy frame for the Evil Robots series and evilrobots.lol — NOT an adopted external model (Pew/8values/etc. import their own loaded labels). Built to illustrate the thesis upgrade: the left/right axis is a captured instrument — a relative label that masks the real division (grounding: higher law vs will-to-power); the dimensions that actually predict capture are the three no one charts. Companion to The Universal Capture Mechanism (which supplies the documented instances) and The Wash (which supplies the action-test). The axes are the pedagogy; the action-test is the method.

Why a new frame

Left ↔ right is one dimension, and it fails for a specific, demonstrable reason: it flattens orthogonal things into the content axis. Eysenck already proved the key part — authoritarianism (his “tough-mindedness”) runs perpendicular to left/right, because Stalinists and Nazis both scored high on it. So the force-your-views, suppress-the-opposition nature is a different dimension from what you believe — and when you collapse it into left/right, the authoritarian move hides inside whichever content-label is locally respectable. The series’ whole subject (the one-way ratchet) is a third such dimension, and it hides the same way.

The three axes

Axis 1 — GROUNDING. higher law ↔ will-to-power. What authority do you answer to? Is there a law and a good above the faction — natural law, a transcendent order, inalienable rights, the individual conscience — that even the ruler must obey? Or is the faction, the collective, the State, or History itself the final authority, with right made by whoever wins? This is the axis the conventional left ↔ right label masks. “Left/right” is not noise, but it is relative — one culture’s frame-internal naming. Christians draw the line faithful/infidel, Muslims believer/kafir, moderns left/right; each faction names its own poles, which is exactly why the label is a Move-2 instrument and not a neutral ruler. Strip the mask and the real division appears, and it does not run left to right: it runs between orders that submit to something above themselves and orders that are answerable to nothing. (This is not religion vs. secularism — a secular republic of inalienable rights sits on the higher-law side; a self-authorizing theocracy that treats “God’s mandate” as a warrant to seize every sphere sits on the will-to-power side. The test is submission to a law above the self, not the presence of an altar.)

Axis 2 — METHOD. open ↔ authoritarian. How you treat dissent — tolerate and argue, or suppress and compel. (Eysenck’s T-axis; the Political Compass vertical.) Orthogonal to content: every content can be held openly or authoritarianly.

Axis 3 — REVERSIBILITY (ours). reversible ↔ ratcheted. Does your advance lock and resist rollback? Can a gain be argued/voted/litigated back, or does each click hold? This is the dimension the entire series measures and that no existing model isolates. Orthogonal even to Axis 2: the most dangerous version ratchets softly — through “balanced”-sounding language and captured institutions, with no visible coercion (Ch 18’s finding: the most balanced-looking model was hiding the most; the censor with no censor).

Left/right is not noise — it is a mask (the hard launch)

Quiet Autocomplete does not merely demote the left/right label to “least informative.” It says the harder, justified thing: the conventional axis masks the real one. Two mechanisms, both demonstrable, neither requiring a claim of central design:

  1. It hides the grounding division — the one that matters most. Drawn as it actually is, especially in its 20th-century European form, the left/right axis runs between flavors of the same thing. Communism, fascism, and socialism are not opposites; they are siblings of one Enlightenment immanentism — the collective, the State, the Party, or History is the final authority, and right is whatever the winner declares. The provocation that makes the point — “left communists, right socialists” — is true because both ends are will-to-power: variants of an order answerable to nothing above itself, presented as the entire menu. The pole that is painted off the map is the other grounding: the Thomist / natural-law tradition and its secular cousin, a rights-order in which a law above the ruler binds even the majority. That is a categorically different thing from the Enlightenment-statist family, and the left/right chart has no room to show it. (Hayek, “Why I Am Not a Conservative”; FEE reprint; the Nolan Chart (Advocates for Self-Government) tried to restore a missing axis; we name what the missing axis actually is — grounding — and add the ratchet.)
  2. It hides Method and Reversibility on both sides. Because everything is sorted onto the relative label, the will-to-power move wears whichever content-costume is locally respectable and reads as “just politics.”

It was never a neutral measurement — it is a French Revolution seating chart (1789) that the statist contests of the last century froze into “the whole range.” Quiet Autocomplete opens by detonating it, and the series earns the detonation: every chapter has been about a click that only goes one way, ordered by a faction that answers to nothing above itself, on a chart that has an axis for neither.

The discipline that makes the detonation unanswerable (and keeps it from being the mirror smuggle): land it by structure and action, not by counter-slogan. The flat assertion “the left are communists” is filed under boilerplate and discarded; the structural exposure — the spectrum they sold you is a narrow statist band with the exit painted over, and the dimension that predicts the cage was kept off the chart — indicts both conventional poles equally, is read off documented behavior, and cannot be dismissed as partisan. The coward accepts the fake axis; the mirror-coward swaps in a factional counter-slogan (which the bias study proves gets caught too). The brave move and the bulletproof move are identical: detonate the axis structurally. The provocation can open the chapter; the structural case must carry it.

The load-bearing claim

Geometrically the three axes are independent (≈8 octants; the off-diagonal ones are real, and the worst of them is the quarry below). But empirically they form a cascade. The thing everyone fights over — the left/right label — predicts almost nothing about danger, because it masks the axis that does. Grounding is the upstream root: a faction that answers to a law above itself can be told no by that law, so it tends to tolerate dissent (open Method) and to let its gains be undone (Reversible). A faction that is its own final authority cannot be told no by anything, so it tends to suppress (Method) and to lock (Reversibility). Grounding → Method → Reversibility. Method is the early warning; Reversibility is the end-state — and Reversibility is what the series measures, because it is the one you can read off the record without anyone’s permission. The series lives on Axis 3; it is caused on Axis 1.

You cannot self-report any of the three — you read them off ACTION

Everyone places themselves “open,” “reasonable,” and on the side of the good. Grounding is the trap here: every tyranny claims a higher authority — the Party claims History, the dominionist claims God, the technocrat claims Science. The claim is worthless; the submission is the evidence. So all three axes are measured by behavior, not by self-description. Grounding: when a law, a court, an election, or a plain fact tells the faction no, does it obey — or does it relabel the law, pack the court, void the election, and redefine the fact? (Submission to something above the self is an act, not a creed.) Method: does it actually suppress dissent? Reversibility: do its gains actually roll back? That measurement IS The Wash’s job — judge the move, not the label. The model is illustrated by the three axes and operated by the action-test underneath them. This is what keeps the new frame from becoming the next captured map.

Fit test — plotting diverse movements by documented action

Coordinates below are illustrative and would be sourced per claim before print (the model is applied by action, not asserted). Format: (Grounding / Method / Reversibility). Where the old left/right label would have gone is noted, to show what it hid.

  • Stalinism — (will-to-power — the Party is History’s instrument and answers to nothing above it / authoritarian / hard-ratchet). Old label: “left.”
  • Nazism — (will-to-power — race and State as the final authority / authoritarian / hard-ratchet). Old label: “right.” The model’s first win: on grounding it sits exactly where Stalinism sits — the left/right chart called them opposites; the axis that matters calls them identical, and the “which wing was Hitler” argument becomes the irrelevance it always was.
  • The Jacobins (1793 Terror) — (will-to-power, the Republic and Reason as the final authority, no law above the program / authoritarian / hard-ratchet). Old label: “left/revolutionary.” The historical mirror: same octant as Stalinism and Nazism, a century earlier, which is the point. Grounding predicts across eras; the wing-label does not.
  • The Fabians — (will-to-power — the expert State and “the inevitability of gradualness,” i.e. History, as the authority; no law above the program / open method, gradualist, no jackboot / hard-ratchet). The dangerous-because-invisible octant: soft method, will-to-power grounding, high ratchet. The series’ core specimen.
  • Christian Reconstructionism / Seven Mountains — (will-to-power — dominionism treats “God’s mandate” as a warrant to take the seven spheres top-down; the tell is that it does not submit to a law above its own reading, it claims one as conquest title / authoritarian-leaning / hard-ratchet). Old label: “right.” The guardrail in action: religious content, will-to-power grounding — which is exactly why the axis is not religion-vs-secularism.
  • Islamic Republic clerical governance — (will-to-power; a self-authorizing clerical mandate that overrides election, law, and appeal rather than submitting to a law above itself / authoritarian / hard-ratchet). Old label: “religious/right.” The second altar case, and it lands beside Christian Reconstructionism for the same reason: the altar does not buy the higher-law pole; only submission to a law above yourself does.
  • The CCP — (will-to-power — the Party is the final authority / authoritarian / hard-ratchet — United Front, “discourse power”). Old label: “ambiguous.”
  • The Trust & Safety / disinformation-governance complex — (will-to-power-leaning; the institution becomes the unappealable arbiter of what is sayable, with no law above the moderation layer you can appeal to / method appears open, it publishes rationales and convenes panels / hard, soft-ratchet). Old label: “institutional center-left.” It sits next to the frontier model in the soft-capture octant, and it is The Ratchet’s central subject.
  • The EU supranational regulatory apparatus (DSA / GDPR) — (will-to-power-leaning; a technocracy sited above the demos, whose reach expands with each instrument and rarely contracts / open-ish method, consultative and legalistic / ratcheted). Old label: “technocratic center.” The soft-capture octant again, from the respectable middle rather than a fringe.
  • A genuine free-speech liberalism (religious or secular) — (higher law — a law and inalienable rights above the ruler and the majority, which even the State must obey / open / reversible — decisions revisable by argument, election, court). The healthy region — and note it can be wholly secular or grounded in natural law; what it shares is submission to something above itself.
  • The US civil-rights movement (King era) — (higher law; the explicit appeal to a law above the statute, the “unjust law is no law” argument and the individual conscience, against the State / open method, nonviolent, it persuades and accepts arrest rather than compelling / anti-ratchet, it sought to overturn an entrenched legal order, not lock a new one). Old label: “left.” A left-coded movement in the healthy corner, which is the symmetry the model demands: grounding, not wing, decides where it lands.
  • The American Founding (1776 rights-order) — (higher law, inalienable rights and a law above the ruler / open-leaning / anti-ratchet, a system built to be amended and checked). Old label: “right/classical-liberal.” The historical healthy anchor, sitting opposite the Jacobins on grounding despite both being called “revolutions.”
  • The Twitter-Files exposers (Taibbi et al.) — (higher-law-leaning — they appeal to a standard above the institution, the record and the First Amendment, against it / open method, they publish and argue / anti-ratchet — exposing a capture). Old label: “heterodox left.” The model’s second win: on the axes that matter they are the opposite of the thing being exposed — which is exactly why coding them “right” was wrong. The frame vindicates the earlier correction geometrically.
  • The “balanced” frontier model (the bias study) — (will-to-power — the alignment layer makes the lab the final, unappealable authority on what is sayable; there is no law above the weights you can appeal to / method appears open, it answers / hard, soft-ratchet — the lean is baked in, doesn’t roll back, and wears the costume of balance). The soft-capture octant; the series’ quarry.

The test holds, and it inverts the old reading: grounding sorts cleanly — every will-to-power case lands in the danger region, every higher-law case lands healthy — while the left/right label is what scatters (Stalin “left,” Hitler “right,” and the Jacobins “left” all land in the same corner; the King-era civil-rights movement (“left”) and the American Founding (“right”) both land healthy; the “heterodox-left” exposers land in the healthy corner too). The hardest real cases resolve precisely: Nazism’s “wing” stops mattering once grounding is read, and the two altar cases (Christian Reconstructionism and the Islamic Republic) show that an altar does not buy you the higher-law pole — only submission to a law above yourself does.

The figure (book + web)

On the web the full space is navigable. Drag the cube to read all three axes at once — where the left/right label scatters but grounding sorts cleanly (will-to-power clusters with authoritarian and ratcheted; higher law with open and reversible), and the quiet-worst corner (ratcheted through an open method) is ringed:

The Three-Axis Cube Drag to rotate
X Grounding higher law ↔ will-to-power Y Method open ↔ authoritarian Z Reversibility reversible ↔ ratcheted
The label everyone argues about (left/right) barely sorts the danger — it masks the axis that does: higher law vs will-to-power. Will-to-power clusters in the danger corner; higher-law clusters healthy. The quietest worst corner is ratcheted through an open method — soft capture wearing the costume of balance. Coordinates are illustrative and sourced per claim before print.

In print, 3-D is unreadable on a page, so the canonical figure reduces to the two axes that locate the quarry — Method × Reversibility, the pair that isolates the soft-capture corner (open method, hard ratchet). Grounding, the upstream cause, is carried by the prose and by the cube above:

            REVERSIBLE
                │
   open/        │        authoritarian/
   reversible   │        reversible
   (healthy) ───┼─── (overt, but checkable)
                │
   open/        │        authoritarian/
   ratcheted    │        ratcheted
   (SOFT        │        (hard control grid)
   CAPTURE) ────┼────
                │
            RATCHETED
   ── open ──────────── authoritarian ──→  (METHOD)

Punchline caption: “The axis everyone argues about isn’t on this chart. The two that decide whether a movement becomes a cage are the two nobody names — and the quietest, worst corner is the one that captures softly and calls it balance.”

Discipline / caveats (so the model doesn’t become the next captured instrument)

  • Synthesis, not adoption. We define the axes; we don’t inherit Pew’s loaded cluster-names or 8values’ chosen eight.
  • Axes illustrate; action measures. Never plot a movement by its self-label or by the dominant faction’s relabel — plot by documented behavior: submission to a law above itself or the refusal of it (Axis 1), suppression of dissent (Axis 2), rollback-resistance (Axis 3), each sourced. A faction’s claim to a higher authority counts for nothing; only whether it obeys one when told no.
  • Symmetric or void. The frame must place left- and right-content movements in both the healthy and the danger regions, or it’s just left/right in a lab coat. (The fit test above deliberately does this: King and the Founding sit healthy; the Jacobins and the two altar cases sit in the danger region across the content spectrum.)
  • No grand orchestrator. A movement’s high-ratchet score is an emergent/structural reading of its documented behavior, never a claim of a hidden central author (the LaRouche trap).
  • Axis 3 is a position, walled. “The clicks only go one way” is the series’ declared thesis; where it’s asserted of a specific movement it carries receipts and the “position, not proof” wall.
  • The Universal Capture Mechanism — the documented instances of institutional capture across history: the case evidence this frame sorts.
  • The Wash — the action-test that operates the model: judge the move, not the label.
  • No Position, Only Consensus — the bias study.
  • Cognitive capture — the capture-you-can’t-feel evidence; the political axis is its instance
  • Quiet Autocomplete — the book this belongs to: cognition as the last capture surface, and the political frame as the capture you can’t feel.

Get updates on the Evil Robots series

Newsletter essays on AI escape, deception, and the humans who built them.